PROCEEDINGS
f the
INNEAN
SOCIETY
of
NEW SOUTH WALES
VOLUME 131
Pro, :
SI ALELIII
; SCS ER
TPS
ee a
NATURAL HISTORY IN ALL ITS BRANCHES
THE LINNEAN SOCIETY OF
NEW SOUTH WALES
ISSN 0370-047X
Founded 1874
Incorporated 1884
The Society exists to promote the cultivation and study
of the science of natural history in all its branches.
The Society awards research grants each year in the
fields of Life Sciences (the Joyce Vickery fund) and
Earth Sciences (the Betty Mayne fund), offers annually
a Linnean Macleay Fellowship for research, contributes
to the stipend of the Linnean Macleay Lecturer in
Microbiology at the University of Sydney, and
publishes the Proceedings. It holds field excursion and
scientific meetings, including the biennial Sir William
Macleay Memorial Lecture delivered by a person
eminent in some branch of natural science.
Membership enquiries should be addressed in the first instance to the Secretary. Candidates for election
to the Society must be recommended by two members. The present annual subscription is $A56.00.
The current subscription rate to the Proceedings is set at A$110.00 per volume. In recent years
a volume consists of a single annual issue.
Back issues of all but a few volumes and parts of the Proceedings are available for purchase. Prices are
listed on our home page and can also be obtained from the Secretary.
OFFICERS AND COUNCIL 2010/2011
President: David Keith
Vice-presidents: M. Cotton, D.R. Murray, M.L. Augee
Treasurer: I.G. Percival
Secretary: J-C. Herremans
Council: M.L. Augee, J.P. Barkas, M. Cotton, M.R. Gray, J-Cl. Herremans, D. Keith, R.J. King, H.A.
Martin, E. May, D.R. Murray, P.J. Myerscough, I.G. Percival, J. Pickett, S. Rose, H.M. Smith and K.L.
Wilson
Editor: M.L. Augee
Assistant Editor: Elizabeth May
Auditors: Phil Williams Carbonara
The postal address of the Society is: P.O. Box 82, Kingsford NSW 2032, Australia
Telephone: (International) 61 2 9662 6196; (Aust) 02 9662 6196
E-mail: linnsoc @iinet.net.au
Home page: http://linneansocietynsw.org.au
Cover motif: Reconstruction of the placoderm fish Sherbonaspis hillsi
Young and Gorter (1981), from Hunt and Young, this volume, page 76.
PROCEEDINGS
of the
LINNEAN
SOCIETY
AUG
of —
NEW SOUTH WALES
For information about the Linnean Society of New South Wales, its publications and
activities, see the Society’s homepage
http://linneansocietynsw.org.au
VOLUME 131
July 2010
vy
i
* i »
Pf f
,
<=
"
if
i i)
7) 1 1 .
W a
: Si : _ i VE
ie
¥
A if vee
wi,
y f
: ,
»
Hye
i ie
a1
! a
\ = t
io 1; t ‘ «i
i
\
=
A a
_ <<
ae ay ‘ i
4
id
Tasha
Sb
: i
“
af
D i}
'
’ ‘ b
{
i i ax
=
: é
i i : Xx
i)
ts .
Mi
i‘
& ® @ =
i
‘
‘
‘
hi
| wire ‘
vy
ry
YT
)
i
j
v
Pal |
.
i
dé
{
1 i r y
:
si) fie
chet
i ip
: 7 i
ee |
i
we ai | i -
\
' i
|
aa
1
i I Pid
4 * ,
. > = A
= 5 us ib r:
( t ow Hye. 't tele Oman
ay i a a med We
~ |) =a eee ee = its
f n . Pik!
, desl Po 991% wah :
ical Me phe ‘ come
4 = = i)
=
, |
7 AL -
2
2
a 2 ri
- a
} aT:
y 7 i
ee. he,
ingen ¥.
ran
~
\ VW
{ ost
he
Sau 7
di A ee an
2 >. ai
2 YD
K
7 7 n “yo ‘ Hy
| if : 3 ? : _
ne +) a ee i 3. i a ;
n , 7” ‘Oe + a
SS a i 1) i a - be ia ,
; Si a = _ —_ )
cS * tnt aa oe - Pet
A oa eG RRS Ore SO ay” 1S. e
7 : ‘ a 7 7 - iY i - es a _
a oe:
de re
_ - nian - , . 2 - - te ire i!
7. veer |
_ 2 af = an oa ¥
IMPORTANT NOTICE TO SUBSCRIBERS AND MEMBERS
Volume 132 will be the last issue of the Proceedings of the Linnean Society of New South
Wales to be published in hardcopy, paper format. Starting with Volume 133, the Proceedings
will be published electronically.
Subscriptions and membership fees paid for 2010 will cover both Volumes 131 and 132.
At the end of each year, beginning December 2011, papers published on the net during the
year will be compiled as the next sequential volume and put onto a CD disc. That disc will
be provided free of charge to all subscribers and members who were financial at the end of
2010.
During the year, as manuscripts are received and processed, they will be put onto the net
immediately when accepted.
This is clearly the most dramatic change since the first volume was printed in 1875. However,
as can clearly be seen from the small size of this volume, authors are moving away from the
slow, traditional means of publication to the more rapid and more widely distributed services
offered by electronic journals.
The Linnean Society of NSW will, starting with Volume 133, offer rapid publication of
fully refereed papers. The net is of course full of unreliable clutter, and the Council of the
Linnean Society of NSW feels that the Society can continue to make a valuable contribution
to natural science in Australia by providing a source of peer-reviewed scientific publications.
To this end, the Proceedings of the Linnean Society of New South Wales will be published on
the escholarship site of the University of Sydney Library.
Eventually the Council of the Society intends to scan all of our back issues and place them on
the escholarship site with open access. At present the Society does not plan to charge users
for access to any of our material.
Full details will be published in Volume 132 and on the Society’s web page.
Michael L. Augee
Editor
in
7
Liftts
. :
7 =>
iT i
p
i »
- a
7 J
M's 2 eRe Gt aoTTO TVATHOUAE mee
Hons tt agli 4b le cone t0i od ad ai
4 wen i tio errede Jee eee vest
joy 4th } Mra Oil tt biae 23) oot (hee
ro =
vie >» *
, ; ~ & 7 L a at
) vray detache ly rapes), EPO ait) gr he a tte dairy WP bers ve
| wf? oath Gly nt! nia Wy gaurd talingupes lyon ab ae bali )
fun, of yy ah ww ie eters Lil, hl ue it 1 eyes a “eft Res)
cee
¥ | ay xaigin J 10) ‘Voc | oon] IWAN guy) Cin. Lire we say ag
; 24 a brs inyrooe raw atsal
S9yS40H 278 mg barn W Sino fev ed! gonis oyna ame or Wann sai
asts srverti RA Ul ran ie wre orrifho v iO WR I Mpernz edt cout (i998 = at yhesio i
" 4 ft ! Srlitiny ba?) Ie Diw Sie nr tye ry Te iP Ogi vile yg "hey znueatrt (wince a ‘
¥ iruhad, anarpalg dk
90 moliesiidug blige 288, £61 onwloy dow one ftege | We lo etn
ai! harivg ) Ste bead oii! } edad | i. tavneici hs ot. toe wht ieee
ouleiAaads sidan & ax bat: n 23)908 oi? dayii. atout Wate ty ys ert
ail nibs ahversioe bans HITS | rte By aqnibi nye ool stacy Ig I . 6
no bulaitdua.od Miw es\ou Wud? Hor (6 2, CASA ibe) taal WS gina oth is i)
a wale oe Pieris yea (2 Yo tiennyiet aa DTIC pee
nn. croc! alg baw naumee Woe wo lo ln eee a iedaeretent —" sto iaane a
aT) mawsky ob cess All #90 1 VI or Np FIA rete. RIGO MRP nite’ pA
| - ee satin ‘Semon a |
! i Z
Sr dee Pa rapsr oe Vio wt wo sa ade i a
a ry
XE taal, ale
The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales,
Australia. Part 8. The Genera Nilssonia, Taeniopteris,
Linguifolium, Gontriglossa and Scoresbya
W.B. KeitH Hoitmes', H.M. ANDERSON? AND J.A. WEBB?
'46 Kurrajong Street, Dorrigo, NSW, 2453, Australia (wbkholmes@hotmail.com).
Hon. Research Fellow, University of New England, Armidale, NSW.
? 46 Kurrajong Street, Dorrigo, NSW, 2453 Australia.
Hon. Palaeobotanist, South African Biodiversity Institute, Pretoria 0001 South Africa.
3Environmental Geoscience Department, La Trobe University, 3086, Victoria.
Holmes, W.B. K., Anderson H.M. and Webb, J.A. (2010). The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales, Australia. Part 8. The Genera Nilssonia, Taeniopteris,
Linguifolium, Gontriglossa and Scoresbya. Proceedings of the Linnean Society of New South Wales 131, 1-26.
Ten taxa of simple leaves in the genera Nilssonia, Taeniopteris, Linguifolium and Gontriglossa and a lobed leaf
in the genus Scoresbya are described from two quarries in the Middle Triassic Nymboida Coal Measures of the
Nymboida sub-Basin in north-eastern New South Wales. The new species Nilssonia dissita and Taeniopteris adunca
are based on previously unpublished material from Queensland together with conspecific material from Nymboida.
An additional four new species from Nymboida are described; Taeniopteris nymboidensis, Linguifolium parvum,
Gontriglossa ligulata and Scoresbya carsburgii.
Manuscript received 1 March 2010, accepted for publication 29 May 2010.
KEYWORDS: Middle Triassic flora, Nymboida Coal Measures, palaeobotany, simple fossil leaves.
INTRODUCTION
This is the eighth paper of a series describing
the early-middle Triassic Nymboida flora. Part 1 of
this series (Holmes 2000) described the Bryophyta
and Sphenophyta, Part 2 (Holmes 2001) the
filicophyta, Part 3 (Holmes 2003) fern-like foliage,
Part 4 (Holmes and Anderson 2005a) the genus
Dicroidium and its fertile organs Umkomasia and
Pteruchus, Part 5 (Holmes and Anderson 2005b)
the genera Lepidopteris, Kurtziana, Rochipteris and
Walkomiopteris, Part 6 (Holmes and Anderson 2007)
the Ginkgophyta and Part 7 (Holmes and Anderson
2008) the Cycadophyta. In this paper the simple leaves
in the genera Nillsonia, Taeniopteris, Linguifolium
and Gontriglossa together with the enigmatic lobed
leaf Scoresbya carsburgii are described.
A description of the Coal Mine and Reserve
Quarries, the source localities of our described material
together with a summary of the geology of the Basin
Creek Formation, the Nymboida Coal Measures and
the Nymboida Sub-Basin were provided in Holmes
(2000).
METHODS
The material described in this paper is based
mainly on collections made by the senior author and
his family from two then-active Nymboida quarries
(Coal Mine Quarry and Reserve Quarry) over a period
of forty years. The specimens noted in Flint and Gould
(1975), Retallack (1977), Retallack et al (1977) and
Webb 1980 were examined in the collections of the
Australian Museum, Sydney, the Department of
Geology and Geophysics of the University of New
England, Armidale and the Queensland Museum,
Brisbane..
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
The University of Queensland PhD thesis on
“Aspects of Palaeontology of Triassic Continental
Sediments in South-East Queensland” by J.A.Webb
(1980) included the descriptive taxonomy of fossils
of simple leaves, similar to those that form the subject
of this paper. In addition to his own extensive field
collections Webb also examined all available and
relevant material in State and private collections.
Descriptive taxonomy in the past has so often been
based on very limited and often fragmentary material.
From Webb’s extensive range of material it was
possible to gain a better understanding of species
boundaries through the natural range of variation
occurring within the fossil populations. On the basis
of floral similarities, the Esk Formation (Toogoolawah
Group) of south-east Queensland and the Nymboida
Coal Measures of north-east New South Wales were
deposited contemporaneously in the Anisian-Ladinian
(Flint and Gould 1977, Rigby 1977). Regrettably most
of Webb’s research was never published. Because of
its relevance to this paper, two new species presented
below are based on his original descriptions and
types with Webb acknowledged as the author.
Taxonomically comparable Nymboida specimens are
illustrated and listed as “Additional Material”.
Since the completion of the research by Webb
(1980) new studies have been published on similar
taxonomic groups from other Gondwana Triassic
floras that are relevant to this paper. Retallack (1980)
reviewed the Middle Triassic Tank Gully flora of
New Zealand and proposed a new combination
for Linguifolium tennison-woodsii; Artabe (1985)
described six Taeniopteris species from Los Menucos
Formation of Argentina; Anderson and Anderson
(1989), in their taxonomic revision of the SouthA frican
Molteno gymnosperms described and extensively
illustrated nine species of Taeniopteris, five species
of Linguifolium and three species of Gontriglossa;
Gnaedinger and Herbst (1998) described three species
of Taeniopteris and three species of Linguifolium from
El Tranquilo Group of Argentina; Gnaedinger and
Herbst (2004a) described ten species of Taeniopteris
from northern Chile, using a statistical analysis of
venation characters; Gnaedinger and Herbst (2004b)
described one Linguifolium sp also from northern
Chile and Herbst et al (2005) listed one Taeniopteris
sp. and two Linguifolium spp from the Lake District
of Chile.
The Nymboida specimens are preserved in
mudstones, siltstones and sandstones as carbonaceous
compressions or impressions in which the gross
morphology is usually well-preserved. However
spores and cuticles have been destroyed by a tectonic
heating event during the Cretaceous Period (Russel
1994). Therefore our identification of taxa is based
only on characters of gross morphology.
The exact stratigraphic horizon or detailed source
of much of our Nymboida specimens is uncertain as
most were collected from fallen blocks during quarry
excavations. The Coal Mine Qua ry has not been
active for some twenty years but the high working
face, although now rather weathered, provides an
excellent exposure of beds that demonstrate the palaeo-
environmental conditions at the time of deposition
and was described by Retallack (1977). In 2006 the
Reserve Quarry was bulldozed into a featureless
bowl — “for restoration and safety purposes” and the
fossiliferous horizons are now hidden.
The Nymboida material described in this paper
has been allocated AMF numbers and is housed in the
palaeontology collections of the Australian Museum,
Sydney.
DESCRIPTIVE TAXONOMY
Without supporting cuticular evidence and lack
of affiliation with any fertile structures for a definite
systematic placement, the leaves described below are
regarded as form genera in Gymnospermae — sedis
incertae. On the basis of preserved cuticle Nilssonia
leaves with haplocheilic stomata have been placed in
the Cycadales and leaves of taeniopterid morphology
may belong in several groups from ferns to cycads.
Anderson and Anderson (2003) placed their Molteno
Taeniopteris species in the Pentoxylales based
on affiliation evidence and similarly they placed
Gontriglossa in the Gnetopsida. The affinities of
Linguifolium remain uncertain although Retallack
(1980) suggested an affiliation with the seeds
Carpolithus mackayi. Scoresbya has been speculated
as being a fern, a seed fern, a member of the
Caytoniales (Taylor and Taylor 2009) or even a pro-
angiosperm (Weber 1995).
Gymnospermae incertae sedis
Genus Nilssonia Brongniart 1825
Type species
Nilssonia brevis Brongniart 1825
Nilssonia is a form genus that includes simple
linear to oblanceolate leaves to irregularly pinnate
leaves. It has a worldwide distribution and ranges
from the Triassic to the Cretaceous. The main gross
distinguishing character of the leaves is the dorsal
attachment of the lamina which completely covers
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
the mid vein. The appearance of this character is often
an artefact of preservation, eg the fossil may be an
impression of the upper or lower leaf surface or an
internal or external cast or mould that often masks
the form and place of attachment of the lateral veins
to the midrib.
The venation pattern of leaves from Gondwana
localities differs somewhat from that of species
described from the northern hemisphere in the more
common bifurcation of the lateral veins and their
straight and parallel course to the margin. Similar
simple leaves in which the lamina does not completely
cover the mid vein and without preserved cuticle are
placed in the form genus Taeniopteris. Where cuticle
information is available, the haplocheilic stomata and
trichomes indicate cycadalean affinities. No cuticle is
preserved on the Nymboida material. Some specimens
in our Nymboida collections can be placed in a
previously unpublished species as described by Webb
(1980). Note this species is attributed to Webb.
Nilssonia dissita J.A.Webb sp. nov.
Figures 1A—C; 2A, B; 7A
Selected synonymy
1917 Taeniopteris crassinervis (Feistmantel)
Walkom, p.38, Pl. 1, fig. 2.
1975 Nilssonia cf. princeps (Oldham and Morris)
Seward; Flint and Gould, p.71.
1980 Nilssonia dissita Webb, p. 87, P1.11, figs 3, 6,
8, Text figs 18 c, d (Unpubl.)
Diagnosis
Large simple leaf 65-150 mm wide; midrib 2.5—
4 mm in width; lamina covers whole of mid-vein;
secondary veins arise from the dorsal surface of a
moderately wide central rib at fairly acute angle, then
curve broadly to run at 80°—90° to margin; individual
veins frequently bifurcate once, usually as they leave
the central rib, occasionally fork a second time;
density of venation 9-16 / 10 mm.
Description (revised to include new Nymboida
material)
Leaves are simple, oblanceolate with undulate to
entire margins and wavy to smooth surface, tapering
to obtuse apex. Length from c. 200 to >300 mm, the
leaf base is not known; width at mid lamina ranges
from 60 —150 mm. Lamina is dorsally attached and
completely covering the mid vein. Lateral veins
diverging from a mid point above the mid vein at
an angle of 50°-70°, arching to run at a high angle
(70° — 90°) straight and parallel to the margin. Many
Proc. Linn. Soc. N.S.W., 131, 2010
veins bifurcate once, usually as they leave the central
rib; a few subsequently fork a second time but never
anastomose; veins coarse with a density 9-16 / 10
mm. Mid vein when exposed ranges in width from
1-4 mm.
Holotype
GSQ F12897
Type Locality
Geological Survey of Queensland Locality
1552, Esk Formation, Toogoolawah Group
Additional material
GSQ12898, Esk Fm. UNEF13443, AMF120989,
AMF 130180, AMF 130181, AMF130182,
AMF 130183, all from Coal Mine Quarry, Nymboida
CM. Also the material listed by Webb (1980), mostly
from the Esk Formation of Queensland.
Name derivation
dissitus — Latin — distant, apart, referring to the
widely spaced venation.
Discussion
Previous material from Nymboida (Flint and
Gould, 1975) was recognised by Webb (1980) as
questionably belonging to this species. From our new
collections specimen AMF 130180 is a block showing
two leaves (Fig. 2B), one almost complete, preserved
in almost three dimensions in white sandstone.
The lamina of the more complete leaf, in places,
completely covers the mid vein as can be seen by
the lateral veins appearing to adjoin in mid lamina.
The incomplete specimens AMF130182 (Fig. 2A)
and AMF130183 both show sections of a leaf with
adjoining lateral vein bases over the mid vein. In other
parts of these leaves and similarly in the full length of
AMF130181 (Fig.1C) the mid vein is exposed as an
artefact of preservation. These leaves are included in
this species based on the form, course and density of
their veins and there being no evidence that the veins
were laterally attached to the margin of the mid vein.
Nilssonia moretonti Walkom 1928
Figure 8A
Synonymy
1928 Nilssonia moretonii Walkom, p. 466, Pl. 25,
WES 25 Bs Te
1980 Nilssonia moretonii Walkom; Webb, P1 10, figs
1, 4, 6, 7.
1989 Taeniopteris moretonii (Walkom) Anderson
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
and Anderson, comb. nov. p. 376, fig. 3; p.547,
figs 5, 6.
Description
A simple strap-shaped leaf with entire or slightly
lobed margins; complete leaf unknown, from 30
— 110 mm wide; lamina covering whole of mid vein;
lateral veins departing from a central line above the
mid vein at an acute angle immediately arching then
proceeding straight and parallel to the margin. Veins
frequently fork on leaving the central rib and again
soon after; density 20 — 35 / 10 mm.
Nymboida Material
Known only froma single specimen, AMF 130184
from Coal Mine Quarry, base and apex missing,
vein density in lower portion of lamina 30 / 10 mm
becoming denser distally, to 40 / 10 mm, straight and
parallel at a high angle across lamina and curving
slightly upwards to the margin.
Discussion
This leaf fragment is placed in N. moretonii on
the basis of the very dense venation and its mid dorsal
attachment to the mid vein.
Anderson and Anderson (1989) transferred
Nilssonia moretonii to the genus Taeniopteris
without additional comment. Under “Intergeneric
comparisons” those authors noted that entire
specimens of Ni/ssonia can hardly be effectively
distinguished from Taeniopteris and did not use
the genus Nilssonia. Many of the leaves placed in
Taeniopteris (see below) show evidence of lateral
attachment of the lamina but towards the dorsal edge
of the mid vein. The degree of the lamina overtopping
of the mid vein makes for a subjective differentiation
between Ni/ssonia and Taeniopteris in the absence of
preserved cuticle.
Genus Taeniopteris Brogniart 1832
Type species
Taeniopteris vittata Brongniart 1832
Taeniopteris is a form genus for simple strap-
shaped leaves with entire lamina and occasionally
forking lateral parallel venation running at a high
angle to a prominent midrib and with unknown cuticle
(Meyen 1987, Taylor and Taylor 1993, Anderson
and Anderson 2003). Numerous species have been
described world-wide from the Upper Carboniferous
to Recent. While this leaf form is diverse and
widespread it rarely occurs in abundance. Many
species have been erected for Gondwana Triassic
material, often based on limited or dubious specimens
that do little to demonstrate the natural variation
within a species. Recent papers on Triassic South
American TJaeniopteris have been useful but some
species appear to be based on very few specimens (eg
for Argentina, Artabe 1985, Gnaedinger and Herbst
1998. For material from Chile, Gnaedinger and Herbst
(2004a) have used a statistical analysis of venation
sequence for ten species of Taeniopteris. Triassic
material from South Africa was described by DuToit
(1927) and very comprehensive collections from the
Molteno Formation by Anderson and Anderson (1989,
2003) who described ten species from 29 assemblages
(localities) and used the “palaeodeme approach” and
illustrated the range of variation in a species. From
Australia there are numerous species in the literature
but most have been based on fragmentary material,
inadequate descriptions and have often been poorly
illustrated. Rarely has the natural range of variation
that may exist in a species been recognised. In our
Nymboida collections taeniopterid leaves comprise c.
3% of numbered specimens. Few leaves, especially the
larger forms, are found complete. Occasional bedding
planes (possible sub—authocthonous assemblages)
show numerous individual leaves resembling a natural
autumnal-like leaf fall. In many specimens the leaf
lamina appears to be dorsally attached to the midrib
but without totally covering it as in Nilssonia.
In our Nymboida collections the majority of
taeniopterid leaves fall within the range of variation as
recognised by Webb (1980) from his examination of
over 170 specimens, mostly from the Esk Formation
for his unpublished species Zaeniopteris adunca which
is here validated using his type specimen and slightly
emended diagnosis. Other rare Nymboida leaves with
clearly distinguishing characters are described as the
new species 7: nymboidensis.
Sterile leaves of the enigmatic fern Ogmos adinus
(Webb 1983, Holmes 2001) may be placed as a form
species of Taeniopteris but are not included here.
Taeniopteris adunca J.A.Webb sp. nov.
Figures 3A—H; 4A—C; 5A—C
Selected synonomy
1892 Taeniopteris sp. indet. Etheridge, p. 374, PI.
16, fig. 4.
1924 Taeniopteris (? Danaeopsis) crassinervis
(Feistmantel) Walkom; Walkom, p. 84, Pl. 18,
fiewSe
1925 Taeniopteris carruthersii, Tenison-Woods;
Walkom, p. 85, text fig. 3.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
1965 Taeniopteris aff. lentriculiforme (Etheridge)
Walkom; Hill et al., PL. T8, Fig. 4.
1975 Taeniopteris aff. lentriculiforme (Etheridge)
Walkom; Flint and Gould, PI. 3, figs 8, 9.
1980 Taeniopteris adunca sp. nov. Webb (unpubl.),
Pl. 23, figs 1-11; text figs 51 a-.
Diagnosis
Strap-shaped leaves, very variable in width; leaf
surface rarely undulate; secondary veins always leave
midrib at moderately acute angle, then quickly arch
away and travel straight and parallel to the margin at
70°—90°; individual veins frequently bifurcate twice
but anastomose very rarely; vein density ranging
from 15 to 25 per 10 mm near the margin.
Description
Leaves elongate, strap-shaped; tapering gradually
and fairly uniformly to a stout petiolate base and
distally to an obtuse to acute rounded apex; very
_ variable in size, from 9—60 mm in width and from 110
mm to >250 mm in length; lamina rarely undulate,
margins entire. Midrib sometimes striate, appearing
as a prominent groove or ridge, 1-2 mm wide in mid
leaf and expanding basally to c. 3 mm. Leaf lamina
attached to the dorsal edge of the mid vein without
overlapping the dorsal surface. Lateral veins always
leave the mid vein at a moderately acute angle
(usually less than 45°) and arch rapidly within | to 2
mm then proceed straight and parallel to the margin
at an angle of c. 75° — 85° and more acutely towards
the apex. Veins fork close to the mid vein and then
once or rarely twice across the lamina. Conjoining of
the veins is rare. Density of the veins varies between
populations and leaf sizes and averages c. 15—25 /10
mm near the margin.
Holotype
UQF 18836
Type locality
G. R. 380 551 Blackbutt 1: 63 360 Sheet, Esk
Formation, Toogoolawah Group, Anisian—Ladinian
Illustrated specimens from Queensland
UQF18836, UQF72601, UQF18830, UQF2103,
UQF72814, UQF72813, UQF72811, UQF21494, see
ign:
Additional material
AMF 130185, AMF130186, AMF130187, AMF-
130188, AMF130189, AMF130190, AMF130191,
AMF130193, AMF130194, AMF130215. All from
Coal Mine Quarry, Nymboida CM.
Proc. Linn. Soc. N.S.W., 131, 2010
Name derivation
aduncus, Latin, bent inward, hooked, referring
to the abrupt curvature of the lateral veins as they
leave the midrib.
Discussion
Based on the detailed study of extensive
collections of fossil plant material mainly from
Queensland, J.A.Webb (1980, unpublished)
differentiated two commonly occurring strap-like
Taeniopteris leaf forms mainly on the basis of the
form of attachment of the lateral veins to the mid
vein. Taeniopteris carruthersii, widespread in the
Upper Triassic assemblages, has lateral veins arising
straight from the midrib at a high angle, sometimes
forking and running at almost right angles to the
leaf margin. In 7’ adunca the leaf lamina is attached
dorsally to the midrib with the lateral veins diverging
from the mid vein at an acute angle, usually forking
close to the base then arching and running straight to
the margin at a high angle. This arching of the veins
close to the mid vein is often obscured through the
form of preservation during fossilization but can be
revealed from close examination. While there are
wide variations within the two species and some
overlapping characters, Webb recognised the two
species as distinct and with stratigraphic implications.
T. carruthersii occurs in the Late Triassic Ipswich
Coal Measures whereas 7’ adunca is found in the
Esk Formation of Queensland and the Basin Creek
Formation of the Nymboida Coal Measures, both
Middle Triassic units.
T. adunca is the most commonly occurring form
of Taeniopteris at Nymboida. On some bedding planes
(see blocks AMF130190, AMF130216, AMF130193
and AMF130194) the leaves form an almost mono-
specific assemblage, probably a seasonal leaf-fall.
Both within and between these assemblages there is
a wide variation in leaf size and shape. 7: adunca is
regarded as a species complex.
Taeniopteris parvilocus Anderson and Anderson
from South Africa (Anderson and Anderson 1989)
and from Chile (Herbst et al. 2005) is similar to T.
adunca in outline and size but differs by the less dense
venation (13/10 mm) that runs almost straight from the
midrib and then arches upwards towards the margin.
See below for comparisons with 7: nymboidensis.
Taeniopteris nymboidensis Holmes and Anderson
sp. nov.
Figures 6 A, B
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Diagnosis
Leaf oblanceolate, to 150 mm long, 30 mm wide;
apex obtuse; lateral veins dorsally attached at acute
angle to strong mid vein, widely spaced at point of
attachment, c. 6/10 mm, arching through half the width
of the lamina and then running straight to margin at c.
65°—70°, bifurcating in an irregular pattern, once near
the base and again across the lamina; vein density in
mid lamina c. 14—18/10 mm.
Description
Leaves simple, entire, oblanceolate to 150 mm
long and from 25—30 mm wide, apex obtuse; strong
mid vein 2 mm wide at mid lamina and tapering
distally; base petiolate to >15 mm long. Lateral veins
attached on dorsal edge of the mid vein, decurrent,
widely spaced at point of attachment, c. 6/10 mm,
arching then running straight and parallel to the
margin at c. 65°—70° in mid lamina but more acute
towards the base and apex. Most veins bifurcate
while arching from the base and usually once again
at irregular distances from the margin. The pattern of
bifurcation is very irregular. Vein density in the mid
lamina c. 14—18/10 mm.
Holotype
AMF 130197
Type locality
Coal Mine Quarry, Nymboida, Basin Creek
Formation, Nymboida Coal Measures.
Other material
AMF 130198, Coal Mine Quarry.
Name derivation
nymboidensis- with reference to the type locality
Discussion
Only two slabs in the collections display this new
species. The holotype is on a block on which are the
remains of seven leaves, four appearing to arise from
a common point but the point of attachment is not
preserved (Fig. 6A). 7: nymboidensis differs from T
adunca by its oblanceolate shape, by the arching of
the lateral veins which continues half way across the
lamina and by the irregular bifurcation of the lateral
veins. In shape and venation pattern 7. nymboidensis
is similar to 7: troncosoi Gnaedinger and Herbst
(2004a) but differs by the less dense venation. 7°
fissiformis Anderson and Anderson (1989) is similar
to T. nymboidensis in vein density (15/10 mm) but
is a much smaller leaf; 77 anavolans Anderson and
Anderson (1989) is similar in shape and size but has
coarser venation of c. 12/10 mm.
Taeniopteris sp A
Figure 7B
Description
Mid portion of a very large leaf >100 mm wide;
mid-vein to 5 mm wide, longitudinally striate; lateral
veins attached to the dorsal edge of the mid vein at
60°—70° and quickly arch and run at c. 80° straight
and parallel to each other across the lamina and
curve slightly upwards towards the margin. Some of
the lateral veins bifurcate close to the mid-vein and
others occasionally fork at varying distances towards
the margin. The vein density is ca 10—12/10 mm.
Material
AMF 130199 Coal Mine Quarry.
Discussion
This fragment differs from 7? adunca and T.
nymboidensis by the larger size and broader mid vein
and from N. dissita by the lateral veins not overtopping
the mid vein. Zaeniopteris sp. A of Anderson and
Anderson (1989) from the Triassic Molteno Formation
of South Africa is a very much larger leaf with a finer
mid rib and lateral veins almost overtopping the mid
vein. Another large leaf from the Molteno Formation,
Taeniopteris homerifolius Anderson and Anderson
(1989) has a venation pattern with veins upcurving
towards the margin similar to 7: sp. A but differs
by the lateral attachment of the lamina to the mid-
vein. Webb (1980 p. 218) described a Taeniopteris
sp. (unpublished) with much larger leaves — to 240
mm wide and lateral veins occasionally anastomosing
which he compared with a leaf from South Africa
described by DuToit (1927) as Taeniopteris lata.
Genus Linguifolium Arber 1913 emend. Retallack
1980
Type species
Linguifolium lilleanum Arber 1913
Linguifolium was erected for simple entire
leaves, linear, spathulate, lanceolate or obovate;
apices sub-acute to rounded; with mid vein persistent
to apex; lateral veins arising at very acute angle to
the mid rib then arching to meet the margin at an
acute angle, forking once and occasionally twice
in the nearer third of their length. The status of the
genus Linguifolium was well-discussed by Retallack
(1980). Linguifolium leaves are extremely rare in the
Nymboida collections.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Linguifolium tennison-woodsii (Jack and
Etheridge 1892) Retallack 1980
Figures 8B, C
Selected synonymy
1892 Angiopteridium tennison-woodsii, Jack and
Etheridge, p. 365
1898 Taeniopteris tennison-woodsii, Shirley, comb.
nov. p. 23, Pl. 9, fig. 2.
1947 Doratophyllum tennison-woodsii, Jones and
deJersey, p.37, Pl. 6, fig. 1.
1980 Linguifolium tennison-woodsii, Retallack,
comb nov. fig. 7 F—H.
1980 Linguifolium tennison-woodsii, Webb, p.172,
Pl. 20, figs 1-4, P1.21, figs 1-15, text fig. 41,
a—p, (unpubl.).
1989 Linguifolium tennison-woodsii, Anderson and
Anderson, p.522, figs 1-3.
1998 Linguifolium tennison-woodsii, Gnaedinger
and Herbst, P1.1, fig. d.
Description
A portion of a small linear leaf with the base
missing, tapering slightly distally to an incomplete
apex. Length preserved 80 mm, width 6 mm. Mid
vein not well defined, lateral veins decurrent on mid
vein, arching across lamina to meet entire margin at
c. 75°, forking once close to mid vein. Vein density in
mid lamina c.12/10 mm.
Material
AMF 130200, Coal Mine Quarry, Basin Creek
Formation, Nymboida Coal Measures.
Discussion
Linguifolium tennison-woodsii differs from most
Linguifolium spp. by its narrow linear form and from
the extremely narrow Linguifolium gracile from the
Molteno of South Africa (Anderson and Anderson
1989) by its more arching and denser veins.
Linguifolium parvum sp. nov. Holmes and
Anderson 2010
Figures 9A—C
Diagnosis
Small spathulate sessile leaves less than 100 mm
long, lateral veins decurrent on striated mid vein,
arching across lamina to meet margin at acute angle,
number of veins forking near base variable, very
occasional veins forking and conjoining. Vein density
8-12/10 mm.
Proc. Linn. Soc. N.S.W., 131, 2010
Description
Leaf spathulate; maximum length 100 mm; width
from 11—20 mm, apex rounded, lamina tapering to
sessile base; midrib with longitudinal striations, width
at base 1.5 mm, contracting in width through length
of the leaf; lateral veins decurrent, arching from mid-
vein across lamina to reach the margin at an angle
of 30°-45°; c. half the veins fork once close to the
mid vein; occasional veins fork in the mid lamina and
conjoin to form a long narrow areole. Density of the
veins at mid lamina ranges from 8 to 12/10 mm.
Holotype
AMF130201
Type locality
Coal Mine Quarry, Basin Creek Formation,
Nymboida Coal Measures.
Other Material
AMF 130202, AMF130203, AMF130204, and
AMF 130207 from Coal Mine Quarry. AMF 130205
and AMF 130206 from Reserve Quarry.
Name derivation
parvum — Latin — small, referring to the small
size of the leaves of this taxon..
Discussion
Linguifolium parvum is similar in form to L.
lilleanum Arber (1913), L. ascium Webb (1980)
and L. patagonicum Gnaedinger and Herbst (1998)
but differs by the short length and by the density
and course of the lateral veins. In the Nymboida
collections these Linguifolium leaves are very rare.
The generic diagnosis of Linguifolium states that the
lateral veins do not anastomose. However on some
specimens of L. parvum very occasional lateral veins
fork and conjoin to form a long narrow areole, hardly
reason to remove it from Linguifolium.
? Linguifolium sp. A
Figures 8D, E
Description
A small spathulate leaf somewhat resembling
in shape L. parvum, is 74 mm long and 14 mm
wide, with base and apex missing. The lateral veins
are sparse, c. 8/10 mm and arch slightly across the
lamina at c. 45° to each terminate at a tooth along
a unique finely serrate margin; occasional veins
forking once between mid vein and mid lamina.
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Material
AMF 130208 and counterpart AMF 130209, Coal
Mine Quarry.
Discussion
This form is based on a single specimen and
its counterpart. It differs from all described species
of Linguifolium by the serrate margin. Jungites
polymorpha from the Molteno Formation (Anderson
and Anderson 1989) has a finely serrate margin but
differs by the dense parallel venation and the variably
entire to pinnate lamina margin.
Genus Gontriglossa Anderson and Anderson 1989
Type species
Gontriglossa verticillata (Thomas 1958)
Anderson and Anderson 1989
The genus Gontriglossa was erected by Anderson
and Anderson (1989) for elliptic, petiolate leaves
with veins attached at an acute angle, arching and
anastomosing towards the margin. Some specimens of
G. verticillata from the Molteno Formation of South
Africa (Anderson and Anderson 1989, 2003) show
stems with well-spaced opposite fascicles of three
leaves. From Nymboida, Holmes (1992) described
some reticulate veined leaves that were identified as
Triassic “G/ossopteris-like leaves”. Those leaves are
here transferred to the genus Gontriglossa. Amongst
the Nymboida material is a specimen showing 10
leaves attached in a whorl or a close spiral (10A, 12A).
To accommodate this form in Gontriglossa requires a
slight emendation of the generic diagnosis to include
the attachment of leaves as either terminal whorls,
close spirals or well-spaced opposite fascicles.
Gontriglossa grandis (Walkom) Holmes and
Anderson comb. nov.
Figures 10A; 12A
Synonymy
1928 Anthrophyopsis grandis Walkom, p. 464, text
fio: 2, Pls 265 fie. S:
1992 ?Glossopteris grandis Holmes, p. 122, Pl. 2,
figs1, 2.
Description
Leaves oblanceolate, to 150 mm long, and to
95 mm wide but usually much smaller, attached as a
terminal whorl or a close spiral, apex rounded acute
to obtuse, tapering basally to a short petiole; midrib
distinct, striate; lateral veins leave the midrib at an
acute angle and for about one third of the width of the
lamina they bifurcate and anastomose to form a wide
elongate mesh with a general inclination of c. 45° to
the midrib; for the remainder of the lamina they form
a narrower elongate mesh inclined at 65°—70° to the
midrib; closer to the midrib the meshes are 1-2 mm
wide, wider in the proximal than the distal part, while
towards the margin they narrow to form 7-8 meshes
per 5 mm of width.
Holotype
UQF1724-5, University of Queensland,
Brisbane from Sheep Station Creek in the Esk Beds.
Other material
AMF 78254-78258, Australian Museum,
Sydney — from Coal Mine Quarry, Nymboida.
Discussion
The Nymboida leaves placed in this species are
much smaller (c. 80 mm long and c. 30 mm wide)
than the holotype specimen but are closely similar in
gross form and the anastomosing venation pattern.
The Nymboida specimens are notable for the whorled
or closely spiral arrangement of the leaves. Individual
leaves of G. verticillata (Thomas) Anderson and
Anderson (2003) are similar in size and venation
pattern to the Nymboida leaves but differ by the known
cuticle and the well-spaced opposite attachment of
fascicles of three leaves to an elongated stem.
Gontriglossa nymboidensis Holmes and Anderson
comb. nov.
Figures 11A, B
Selected Synonymy .
1975 Anthrophyopsis grandis Walkom, Flint and
Gould, Pl. 1, fig. 9.
1992 ?Glossopteris nvmboidensis Holmes, P. 122,
Pl. 1p esiSt4 Ply 2ifieele
Holotype
UNEF13528 and paratype UNEF13639, both
from Coal Mine Quarry. Now housed in the Australian
Museum as specimens AMF 126731 and AMF 126730
respectively.
Additional material
AMF 130214, Coal Mine Quarry.
Description
A reticulate veined leaf known only from apical
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
and mid lamina fragments. Leaf of unknown length,
width 50 mm, tapering distally to an acutely rounded
apex; midrib distinct, striated; lateral veins leaving
midrib at c. 20°-30° at intervals of ca 0.5 mm and
quickly arch over a distance of c. 5 mm where they
bifurcate and then run straight to the margin at an
angle of 75°. After the initial bifurcation the veins
fork again two or three times to join with adjacent
veins to form long narrow meshes, each subsequent
mesh being narrower than the proceeding one. The
density of the veins in the mid lamina is c. 12—14/ 10
mm and at the margin c. 18/10 mm.
Additional material
AMF130214, Coal Mine Quarry.
Discussion
G. nymboidensis differs from all other
Gontriglossa species by the very fine narrow parallel
meshes formed by the lateral veins. Cetiglossa balaena
_ Anderson and Anderson (2003) from the Molteno of
South Africa is much larger leaf with more elongate
reticulate venation that does not arch from the mid
vein. The somewhat similar reticulate veined leaf
from Patagonia, Santacruzia hunickenii Gnaedinger
and Herbst (1998) differs by the serrate to incised
margins and the lateral veins attached at a high angle
and running straight to the margin. (See comparison
of Santacruzia hunickenii with Gontriglossa lacerata
below).
Gontriglossa lacerata (Holmes 1992) Holmes and
Anderson comb. nov.
Figures 11C, D
Synonymy
1992 ?Glossopteris lacerata Holmes, p. 124, Pl. 2,4.
Holotype
AMF78259. Coal Mine Quarry, Basin Creek
Formation, Nymboida Coal Measures.
Additional material
AMEF130210 and AMF130213 from Reserve
Quarry
Description
Known from three incomplete specimens. Leaf
broad-elliptic or oblanceolate, >180 mm long, 65
mm wide, petiolate; apex broadly rounded; margin
iregularly lacerate, dentate or lobed; venation
somewhat similar to G. nymboidensis, arching
from mid-vein, bifurcating and anastomosing to the
margin.
Proc. Linn. Soc. N.S.W., 131, 2010
Discussion
This is a bizarre species. It differs from other
Gontriglossa species by the irregularly lacerate
margins which we believe to be natural and not
resulting from insect damage.
Gnaedinger and Herbst (1998) described
from the Triassic Tranquilo Group of Santa Cruz,
Argentina a leaf with reticulate venation and serrate
to deeply incised margins and placed it in their new
genus and species Santacruzia hunickenii. They were
perhaps unaware of the paper by Holmes (1992) as
they made no comparisons with ?G/ossopteris (now
Gontriglossa) lacerata. S. hunickenii differs from
Gontriglossa retculata by the less deeply incised
margin and by the much denser venation that passes
at 90° from the mid-vein to the margin. Gnaedinger
and Herbst did compare Santacruzia with the
Molteno species Gontriglossa balaena that has been
transferred to the genus Cefiglossa Anderson and
Anderson (2003) which lacks the lacerate lamina
margin.
Gontriglossa ligulata Holmes and Anderson sp.
nov.
Figures 12B—D
Diagnosis
Leaf ligulate, lateral veins decurrent on mid vein,
widely spaced, arching and bifurcating once then
running straight at a high angle towards the margin;
forking again in mid lamina and conjoining to form
a longitudinal row of transverse rhomboidal areoles
and a row of triangular areoles parallel and adjacent
to the margin.
Description
An incomplete strap-shaped leaf 80 mm long
but with base and apex missing; lamina 14 mm wide
above broken base, tapering gradually over whole
length to 8 mm; mid vein 1 mm wide; lateral veins
decurrent and widely spaced on mid vein, arching
and bifurcating once then passing to margin at c. 75°.
Between mid lamina and margin each vein bifurcates
twice and anastomoses with adjacent veins to form
a longitudinal row of transverse rhomboidal areoles
and a row of triangular areoles parallel to the margin;
vein density near margin c. 16/10 mm.
Holotype
AMF 130211
Type Locality
Reserve Quarry, Nymboida, Basin Creek
Formation, Nymboida Coal Measures.
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Name derivation
ligulata — Latin, strap-shaped, referring to the
broad-linear form of the leaf.
Discussion.
This new species is based on a single incomplete
specimen. While recognising that some species
of Taeniopteris, eg T. fissiformis and T: anavolans
(Anderson and Anderson 1989; Gnaedinger
and Herbst 2004a) may show rare and irregular
anastomoses, we believe that from the regular and
distinctive anastomosing venation (see Fig. 12D)
this leaf is best placed in Gontriglossa, The linear
shape of the leaf and the details of the anastomosing
venation pattern differentiate G. /igulata from the
other Gontriglossa species described above and from
the cordate based leaf, G. hilaryjanea (Anderson
and Anderson 1989, 2003). The regular form of the
marginal areoles diffentiates G. /igulata from the
Scoresbya sp. described below.
Genus Scoresbya Harris 1932
Type species
Scoresbya dentata Harris 1932
Scoresbya dentata was described by Harris (1932)
for small palmate leaves with reticulate venation and
dentate margins from Scoresby Sound in the Jurassic
of Greenland. Additional specimens of Scoresbya
dentata have been described from the Jurassic of
Germany (Krausel and Schaarschmidt 1968), from
China (Cao 1982), Afghanistan and Iran (Schweitzer
and Kirchner 1998) plus an additional species
from the Late Triassic of Mexico (Weber 1995).
An incomplete specimen showing parts of several
segments of a palmate leaf with dentate margin and
reticulate venation from the Ipswich Coal Measures
of Queensland was described by Shirley (1898) as
Phlebopteris (?) dichotoma and later transferred by
Herbst (1974) to the Scoresbya genus.
Scoresbya carsburgii Holmes and Anderson sp.
nov.
Figures 13A, 14A, B.
Diagnosis
A large leaf bifurcating irregularly into broad
linear lobes; margins entire to irregularly serrulate;
lateral veins decurrent on striate mid vein, then
arching and running to margin, forking near base,
occasionally in mid lamina and then forking and
10
sometimes conjoining to form small areoles adjacent
to the margin; vein density in mid lamina c. 12 / 10
and c. 18 / 10 mm near margin.
Description
An incomplete palmate leaf; mid _ vein
longitudinally striated, 3 mm wide in proximal
section of leaf; lamina bifurcating at 10 mm from the
base of leaf as preserved. The minor fork produces a
broad linear pinna or lobe 90 mm long and 28 mm
wide. After 43 mm the main rachis again bifurcates
to form a major elongate lobe (pinna) 120 mm long
and 30 mm wide and a minor lobe 60 mm long and
20 mm wide, both tapering slightly distally. The
margins of the lobes are entire to irregularly undulate
or serrulate. Throughout the leaf the decurrent lateral
veins are widely spaced as they arch at an acute angle
from the main rachis, soon forking irregularly and
then running straight to the margin at c. 30°-45°,
again sometimes forking at irregular distances across
the lamina; close to the margin some veins again fork
and conjoin to form small triangular areoles adjacent
and parallel to the margin (Fig. 14B). Density of the
lateral veins in mid lamina c 12/10 mm and near the
margin c 18/10 mm.
Holotype
AMF130212
Type Locality
Reserve Quarry, Nymboida, Basin Creek
Formation, Nymboida Coal Measures.
Name derivation
carsburgii — named for the collector of the
specimen, amateur fossil plant and insect enthusiast,
Mr Allan Carsburg.
Discussion
Scoresbya carsburgii is based on a single
incomplete specimen that overlies another lobe
fragment. It differs from the northern hemisphere
species S. dentata Harris by its larger size, less
obvious dentate or pinnatifid margins and by the
form of venation. Scoresbya dichotoma (Shirley)
Herbst (1974) from the Ipswich Coal Measures of
Queensland is a smaller leaf and as described by
Herbst has veins conjoining to form an intramarginal
vein similar to that in the genus Yabiella. From the
late Triassic of Chile Mollesia melandeziae Melchior
and Herbst (2000) is described as particularly
similar to Scoresbya but with a different venation
pattern. The affinities of Scoresbya are not well
understood. Herbst (1992) excluded Scoresbya from
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
the Dipteridaceae and Taylor et al (2009) discussed
it under the Caytoniales while Weber (1995) inferred
a possible link with angiosperms. S. carsburgii is
an interesting addition to the Nymboida flora and
illustrates the many puzzles still to be solved in these
ancient floras.
CONCLUSION
This paper deals with leaves of simple form
placed in the form genera Nilssonia, Taeniopteris,
Linguifolium and Gontriglossa and a unique lobed
leaf referred to the genus Scoresbya. Described are
two species of Nilssonia including a new species N.
dissita; three species of Taeniopteris including the
new species 7. adunca and T: nymboidensis; two
species of Linguifolium including the new species L.
parvum; four species of Gontriglossa including three
new combinations and a new species G. /igulata.
_ A unique specimen of a lobate leaf is described as
Scoresbya carsburgii sp. nov.
ACKNOWLEDGEMENTS
WBKH deeply appreciates the assistance
provided by his daughters Marnie and Netta and
late wife Felicity in collecting from the Nymboida
localities over many years. Drs Susan Parfrey and
Kristen Spring of the Queensland Museum Collections
kindly located specimens described in Webb’s Thesis.
WBKH is assisted by a grant from the Betty Main
Research Fund.
REFERENCES
Anderson, J.M and Anderson, H.M. (1989). Palaeofiora of
southern Africa. Molteno Formation (Triassic). Vol.2:
Gymnosperms (excluding Dicroidium). Balkema,
Rotterdam.
Anderson, J.M and Anderson, H.M. (2003). Heyday of the
gymnosperms: systematics and biodiversity of the
Late Triassic Molteno fructifications. Strelitzia 15,
1-398.
Arber, E., 1913. A preliminary note on the fossil plants of
the Mt. Potts Beds, New Zealand, collected by Mr.
D.G. Lillie, biologist to Captain Scott’s Anartctic
Expedition in the “Terra Nova”. Proceedings of the
Geological Society of London B 86, 344-347.
Artabe, A.E. (1985). Estudio systematico de la tafoflora
Triasica de Los Menucos, provincial de Rio Negro,
Argentina. Parte 2. Cycadophyta, Ginkgophyta y
Coniferophyta. Ameghiniana 22, 159-180.
Proc. Linn. Soc. N.S.W., 131, 2010
Cao, Z.Y., 1982. On the occurrence of Scoresby from
Jiangsu and Weichselia from Zhejiang. Acta
Palaeontologica Sinica 21, 344-348.
Du Toit, A.L. (1927). The fossil flora of the Upper Karoo
Beds. Annals of the South African Museum 22,
289-420.
Etheridge, R., 1892. Palaeontology: in Jack, R.L. and
Etheridge, R., The geology and palaeontology
of Queensland and New Guinea. Queensland
Department of Mines. Geological Survey of
Queensland Publication 92, 1-768.
Flint, J.C.E.,and Gould, R.E. (1975). A note on the fossil
megafloras of the Nymboida and Red Cliff Coal
Measures, southern Clarence-Moreton Basin. Journal
and Proceedings of the Royal Society of NSW 108,
70-74.
Gnaedinger, S. and Herbst, R., (1998). La flora triasica
del Grupo el Tranquilo, Provincia de Santa Criz,
Patagonia. Parte 5, Pteridophylla. Ameghiniana 35,
53-65.
Gnaedinger, S. and Herbst, R., (2004a). Pteridophylla
del Triasico del Norte Chico de Chile. 1. El género
Taeniopteris Brongniart. Ameghiniana 41, 91-110.
Gnaedinger, S. and Herbst, R., (2004b). Pteridophylla
del Tridsico del Norte Chico de Chile.2. Generos
Dejerseya Herbst, Linguifolium (Arber) Retallack y
Yabiella Oishi. Rev. Mus. Argentino Cienc. Nat. n.s.
6(1): 49-59.
Harris, T.M., (1932). The fossil flora of Scoresby Sound,
East Greenland. 2. Meddelelser om Gronland 85,
1-112.
Herbst, R. 1974 Notes on Two Triassic Plants from
Queensland, Australia Proceedings of the Royal
Society of Queensland, 85, 79-84.
Herbst, R. 1992. Propuesta de classification de las
Dipteridaceae (Filicales), con un atlas delas especies
Argentinas. Dorbignyana 6, 1—71.
Herbst, R. and Troncoso, A. (2000). Las Cycadophyta del
Triasico de las Formaciones La Ternera y El Puquén
(Chile). Ameghiniana 37(3), 283-292.
Herbst, R., Troncoso, A. and Mufioz, J., (2005). Las
tafofloras triasicas de la region de los Lagos, Xma
Region, Chile. Ameghiniana 42, 377-394.
Hill, A., Playford, G. and Woods, J.T. (1965).
Triassic Fossils of Queensland. Queensland
Palaeontographical Society, Brisbane. 1-32.
Holmes, W.B.K. (1992). Glossopteris-like leaves from
the Triassic of eastern Australia. In: Venkatachala,
B.S., Jain, K.P. and Awasthi, N. Eds. Proceedings
of the ‘Birbal Sahni Centenary Palaeobotanical
Conference’, Geophytology 22, 119-125.
Holmes, W.B.K. (2000). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 1. Bryophyta, Sphenophyta.
Proceedings of the Linnean Society of NSW 122,
43-68.
Holmes, W.B.K. (2001). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 2. Filicophyta. Proceedings of
the Linnean Society of NSW 123, 39-87.
11
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Holmes, W.B.K. (2003). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 3. Fern-like foliage.
Proceedings of the Linnean Society of NSW 124,
53-108.
Holmes, W.B.K. and Anderson, H.M. (2005Sa). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 4.
Dicroidium. Proceedings of the Linnean Society of
NSW 126, 1-37.
Holmes, W.B.K. and Anderson, H.M. (2005b). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 5.
The Genera Lepidopteris, Kurtziana, Rochipteris and
Walkomiopteris. Proceedings of the Linnean Society
of NSW 126, 39-79.
Holmes, W.B.K. and Anderson, H.M. (2007). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 6.
Ginkgophyta. Proceedings of the Linnean Society of
NSW 128, 155-200.
Holmes, W.B.K. and Anderson, H.M., (2008). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 8.
Cycadophyta. Proceedings of the Linnean Society of
NSW 129, 113-140.
Jack, R.L. and Etheridge, R. Jnr., (1892). The geology
and palaeontology of Queensland and New Guinea.
Queensland Department of Mines. Geological Survey
of Queensland Publication 92, 768 pp.
Jones, O.A. and De Jersey, N.J. (1947). The flora of the
Ipswich Coal Measures —morphology and floral
succession. Papers of the Department of Geology,
University of Queensland. New Series 3, 1-88.
Krausel, R. and Schaarschmidt, F., (1968). Scoresbya
Harris (Dipteridaceae?) aus dem Unteren Jura von
Sassendorf. Palaeontographica 123B, 124-131.
Melchior R.N. and Herbst, R., (2000). Sedimentology of
the El Puquén Formation (Upper Triassic, Central
Chile) and the new plant Mollesia melendeziae
gen. et sp. nov. (pteridophylla, incertae sedis).
Ameghiniana 37, 477-485.
Meyen, S.V., (1987). Fundamentals of Palaeobotany.
Chapman and Hall, New York.
Ottone, E.G. (2006). Plantas triasicas del Grupo
Rincon Blanco, Provincia de San Juan, Argentina.
Ameghiniana 43, 477-486.
Retallack, G.J. (1977). Reconstructing Triassic vegetation
of eastern Australia: a new approach for the
biostratigraphy of Gondwanaland. Alcheringa 1,
247-278. Alcheringa-fiche 1, G1—J16.
Retallack, G.J., Gould, R.E. and Runnegar, B. (1977).
Isotopic dating of a middle Triassic megafossil flora
from near Nymboida, north-eastern New South
Wales. Proceedings of the Linnean Society of NSW
101, 77-113.
Retallack, G.J., (1980). Middle Triassic megafossil plants
and trace fossils from Tank Gully, Canterbury,
New Zealand. Journal of the Royal Society of New
Zealand. 10, 31-63.
Rigby, J.F. (1977). New collections of plants from the Esk
Formation, south-eastern Queensland. Queensland
Government Mining Journal 78, 320-325.
Russel, N.J., (1994). A palaeothermal study of the
Clarence-Moreton Basin. Australian Geological
Survey Organisation Bulletin 241, 237-276.
Schweitzer, H.J. and Kirchner, M., 1998. Die rhato-
jurassischen Floren des Iran, Afghanistan. 11.
Ptreidophyta und Cycadophyta. 1. Cycadales.
Palaeontographica 248B, 1-85
Shirley, J. (1898). Additions to the fossil flora of
Queensland. Queensland Geological Survey Bulletin
7, 19-25.
Taylor, T.N. and Taylor, E.L. (1993). The biology and
evolution of fossil plants. Prentice Hall, New Jersey.
Taylor, T.N., Taylor, E.L. and Krings, M., (2009)
Palaeobotany: The Biology and Evolution of Fossil
Plants. Academic Press. Burlington MA.
Walkom, A.B. (1917). Mesozoic floras of Queensland.
Part | (contd.) The flora of the Ipswich and
Walloon Series. (d) Ginkgoales, (e) Cycadophyta,
(f) Coniferales. Queensland Geological Survey
Publications 259, 1-49
Walkom A.B. (1924). On fossil plants from Bellevue, near
Esk. Memoirs of the Queensland Museum 8, 77-92.
Walkom A.B. (1925). Notes on some Tasmanian Mesozoic
plants.Part 1. Papers and Proceedings of the Royal
Society f Tasmania 1924, 73-89.
Walkom A.B. (1928). Fossil plants from the Esk district,
Queensland. Proceedings of the Linnean Society of
NSW 53, 458-468.
Webb, J.A. (1980). Aspects of the palaeontology of
Triassic continental sediments in South-East
Queensland. Unpublished Thesis. Geology
Department, University of Queensland.
Webb, J.A., 1983. A new plant genus, possibly a
Marattealean fern from the Middle Triassic of eastern
Australia. Memoir of the Association of Australasian
Palaeontologists 1, 363-371.
Weber, R., 1995. A new species of Scoresbya Harris and
Sonoraphyllum gen. nov. (Plantae incertae sedis)
from the Late Triassic of Sonora, Mexico. Revista
Mexicana de Ciencias Geologicas 12, 94—107.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 1. A-C. Nilssonia dissita Webb sp. nov. A. GSQF12897, Holotype, GSQ Locality 1552, Esk
Fm. B. GSQF12898, GSQ Locality 1552, Esk Fm. C. AMF130181 Coal Mine Quarry, Nymboida
CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
13
PRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
14
Figure 2. A. B. Nilssonia dissita Webb sp. nov. A. AMF130182, Coal Mine Quarry. Scale bar =
5 cm. B. AMF130180, Coal Mine Quarry, Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 3. A-H. Taeniopteris adunca Webb sp. nov. A. UQF18836, Holotype. 380 551 Blackbutt
Sheet. B. UQF72601, UQL4110. C. UQF18830, 445 486 Blackbutt Sheet. D. UQF2103. UQL4238. E.
UQF72814, UQL4255. F. UQF72813, UQL4238. G. UQF72811, UQL4110. H. UQF21494, UQLS585.
All from Esk Fm. Scale bar = 1 cm
Proc. Linn. Soc. N.S.W., 131, 2010
15
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Figure 4. A-C. Taeniopteris adunca Webb sp. nov. AMF 130194, Reserve Quarry. B. AMF130195,
Coal Mine Quarry. C. AMF130186, Coal Mine Quarry. All Nymboida CM. Scale bar A, C = 1 cm, B
=S'cmi:
16 Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 5. A— C. Taeniopteris adunca Webb sp. nov. A. AMF130187. B. AMF130189. C.
AMF130196, all from Coal Mine Quarry. Nymboida CM. Scale bar A, B= 1 cm. C =5 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
17
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Figure 6. A, B. Taeniopteris nymboidensis Holmes and Anderson sp. nov. A. AMF130197. B.
AMF130198, both from Coal Mine Quarry. Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 7. A, B. Nilssonia dissita Webb sp. noy. AMF120939. B.Taeniopteris sp A. AMF
130199, both Coal Mine Quarry. Nymboida CM. Scale bar = 1 cm.
Roe. Linn Soe N.S WwW. als 1, 200
19
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
Figure 8. A. Nilssonia moretonii AMF130184. B, C. Linguifolium tennison-woodsii AMF130200. D,
E. Linguifolium sp A AMF130208. Numboida CM. Scale bar A, C, E = 1 em, B=5 cm.
20 Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 9. A— E. Linguifolium parvum Holmes and Anderson sp. nov. A, B. Holotype AMF130201,
Coal Mine Quarry. C, D. AMF130207, Coal Mine Quarry. E,. AMF130206, Reserve Quarry. Nym-
boida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
21
TRIASSIC GYMNOSPERMAE FROM NY MBOIDA - SEDIS INCERTAE
Nw
NM
Figure 10. A. Gontriglossa grandis (Walkom) Holmes and Anderson comb. nov. Holotype AMF
78254 Coal Mine Quarry. Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 11. A, B. Gontriglossa nymboidensis (Holmes) Holmes and Anderson comb. nov. A. Holotype
AMF126730. Coal Mine Quarry. B. Paratype AMF126731. Coal Mine Quarry. C, D. Gontriglossa
lacerata (Holmes) Holmes and Anderson comb. nov. C. Holotype AMF78259 Coal Mine Quarry.. D.
AMF130210, Reserve Quarry. Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010 ™®
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
24
Figure 12. A. Gontriglossa grandis (Walkom) Holmes and Anderson comb. nov. AMF78254 Coal
Mine Quarry. B —D. Gontriglossa ligulata Holmes and Anderson sp. nov. AMF130211, Reserve
Quarry. Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
W.B.K. HOLMES, H.M. ANDERSON AND J.A. WEBB
Figure 13. A. Scoresbya carsburgii Holmes and Anderson sp. nov. Holotype AMF130212,
Reserve Quarry. Nymboida CM. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
DS
TRIASSIC GYMNOSPERMAE FROM NYMBOIDA - SEDIS INCERTAE
26
Figure 14. A, B. Scoresbya carsburgii Holmes and Anderson sp. nov. A. Line drawing of Holotype.
AMF130212. B. Details of venation. Scale bar = 1 cm.
Proc. Linn. Soc. N.S.W., 131, 2010
Catalogue of Insects Collected by William Sharp Macleay in
Cuba 1825-1836
Dominic Cross!. AND ELIZABETH JEFFERYS! .
'The University of Sydney, Faculty of Agriculture, Food and Natural Resource, NSW 2006
(dero3102@uni.sydney.edu.au)
*EAJ Consultants” Principal, 14 Holloway Street, Pagewood NSW 2035
(liz@eaj.com.au)
Cross, D. and Jefferys, E. (2010). Catalogue of insects collected by William Sharp Macleay in Cuba
1825-1836, Proceedings of the Linnean Society of New South Wales 131, 27-35.
All of William Sharp Macleay’s labelled Cuban insects are now in a separately labelled Cuban insect
cabinet in the Macleay Museum. There are over 7,349 labelled, pinned and partially identified. Other
unlabelled specimens are still to be found throughout the collection. The geographical area where Cuba
lies is also within the bio-geographical area for the southern United States, the Bahamas, the Caribbean
and the northern most areas of South America. The biological scientists of these surrounding countries
will find the information and knowledge of the distributions of insects of Cuba found in 1825 to 1836 of
tremendous interest in relation to the possible distributions of insect faunas found or no longer found in
these areas today.
Manuscript received | March 2010, accepted for publication 24 May 2010.
KEYWORDS: Catalogue, Coleoptera, Cuba, Cuban insects, Curculionidae, Havana, Hymenoptera,
Lepidoptera, Macleay Museum, Slave trade, William Sharp Macleay.
INTRODUCTION
The following is a catalogue of Cuban insects
collected by William Sharp Macleay during his
appointment as commissioner for the abolishment
of the slave trade in Havana from 1825 to 1836. The
specimens were taken from Cuba to England at the
conclusion of his posting and consequently were
moved to Sydney, Australia, with W.S. Macleay when
he moved there to live. The collection of over 7000
insects were spread throughout the Macleay Museum’s
entomology collection but were readily identified
using locality labels. This is the first account of the
Macleay Cuban collection and although initially the
collection may have been larger, it is probable that
over 170 years, specimens have had labels removed,
been damaged beyond usefulness, or removed from
the Macleay Museum altogether. All the remaining
labelled Cuban specimens are now reunited in a
single collection and are for the most part in good
condition.
The collection consists of 7349 insects across at
least 11 orders as follows:
Blattodea 33
Coleoptera 2172
Diptera 385
Hemiptera 12D
Hymenoptera 3509
Lepidoptera 407
Neuroptera 40
Odonata 24
Orthoptera ]
Phasmatodea 20
Siphonaptera 29
While care was taken to provide the most up to
date species names, information was not able to be
found on some of the labelled species name, and these
have been included as written on the label. Where the
year has been omitted it is where we were unable to
find the complete documentation of the description
and the publication.
William Sharp Macleay left England for Cuba in
October 1825, to take up his duties in connection with
the Mixed British and Spanish Court of Commission
for the Abolition of the Slave Trade established at
THE MACLEAY COLLECTION OF CUBAN BEETLES
Havana. His residence in Cuba lasted from December
1825 to early in the year 1836. The catalogue of
insects included in this paper, includes all those
insects (over 7000) that are clearly labelled with the
locality Cuba. William collected many specimens
during those eleven years in Cuba, and then brought
them to Australia. All of William Sharp’s collection is
now housed in the Macleay Museum at the University
of Sydney. There may be many more Cuban insects
in the Macleay Museum but this catalogue only deals
with specimens with the label Cuba.
William Sharp Macleay was born in London, on
21“ July 1792, the eldest son of Alexander Macleay
(1767 - 1848) who amassed probably the finest insect
collection in Europe and which eventually Alexander
brought with him to Australia in 1826. William Sharp
Macleay arrived in Australia in 1839 with his own
insect collection from European collecting trips, his
collection from Cuba and a collection of insects from
his trip to the United States with Mr Titian Peale
(Fletcher 1920).
William was educated at Westminster and Trinity
College Cambridge and graduated with a BA in
1814 and MA in 1818. On leaving the University he
was appointed as Attache to the British Embassy in
France. What awakened and developed his interest
in Zoology seems primarily to have been his father’s
example, influence and fine collection of insects.
During his time in Paris he had the opportunity of
meeting Cuvier, Latreille and other distinguished
naturalists of that time, as well as appreciating the
importance of the magnificent establishment of the
Jardin des Plantes. He subsequently was appointed
Secretary to the Board for liquidating British claims
on the French Government, established at the peace
of 1815 - 1825. He was then sent as Commissioner of
Arbitration of the Slave Trade established at Havana
in Cuba. In 1830 he became the Commissary Judge
of the same court. In 1836, he was appointed to be
the Judge of the mixed British and Spanish Court of
Justice, established under the treaty of 1835 — 1836.
In 1836 he returned to England. In 1837 he retired
from the Public Service. He left England in 1838
for Australia with his cousins William and John and
arrived in Sydney in March 1839. Here he continued
to collect insects and studied marine life. He was
also a trustee of the Australian Museum from 1853
- 1862. He was universally recognised as the leading
zoologist in Sydney from 1839 up to the time of
his death. William Sharp died in Sydney on the 26"
January 1865 and was buried in the family tomb in
Camperdown Cemetery (Fletcher 1920).
William Sharp’s published work began in 1819
and ended in 1847 (over 30 published papers). There
28
were no publications on any of the insects that he
collected in Cuba.
During his voyage to Cuba, in the months of
October, November and December of 1825, he made
notes on the Ornithology of the Islands of Madeira,
Teneriffe and Saint Jago, as well as observations
at Barbados, Martinique and off the coast of Saint
Domingo. He always seemed to be taking notes of
his natural surroundings wherever he went. However
there seems to be no detailed notes of his insect
collecting in Cuba, or at least none that has been found.
However there is one interesting letter he wrote to his
trusted friend Kirby, dated 3" 1827 January, about a
year after his arrival. William writes:
“The climate has, I thank God, hitherto agreed
with me much better than that of England: but
there is a languor attendant upon every kind of
exertion, which makes reading or study here a
very different thing from what it is in England.”
“This is a good place for Wading Birds, Lizards,
Butterflies and Sphinges, (a term meaning Hawk
Moths ), but apparently nothing else. I live in
the country, where I have a large house and
garden: this is my principal amusement, as I
take great pleasure in cultivating Orchideae,
particularly those which are parasitical on trees.
The disagreeable are ants, scorpions, mygales
and mosquitoes. The latter were quite a pest on
my first arrival within the tropics, but now I mind
them as much as I did gnats in England. “
The place of his residence in Cuba was
Guanabacoa, (an Indian name meaning “site of the
waters “) which he described as if “ living in the
country is a picturesquely situated amid woods, on
high hills which furnish a fine view, is a town a few
kilometres from the capitol of Cuba, Havana.”
During his leisure hours, natural history soon
began to claim his attention as he sent specimens of
lizards, bats and 45 species of birds to England to be
exhibited at meetings of the Zoological Club of the
Linnean Society in 1828. Later William, sent a foetal
specimen of a dolphin (Fletcher 1920).
While no papers dealing especially with Cuban
insects were published by W.S. Macleay, among his
papers were thirty nine water-colour drawings of
lepidopterous larvae, from which he may have reared
adults. Besides these there are a number of pencil or
pen and ink sketches of Lepidoptera, scorpions, ticks
and mites (Fletcher 1920).
The scientific world of today has been given
an opportunity to know what was on the Island of
Cuba in the years 1825 to 1836 due to the scientific
Proc. Linn. Soc. N.S.W., 131, 2010
D. CROSS AND E. JEFFERYS
endeavours of William Sharp Macleay in the form
of over 7000 dry pinned labelled insects now placed
together as the Cuban insect collection are housed in
the insect collection in the Macleay Museum at the
University of Sydney.
ACKNOWLEDGEMENTS
In July 2009 Dominic Cross was awarded the Macleay
Miklouho-Maclay Fellowship at the Macleay Museum. At
this time his supervisor of the Fellowship was Ms Elizabeth
Jefferys, who was the Curator of natural History at the
Macleay Museum at the University of Sydney. We thank
the Macleay Museum for giving us the opportunity to
complete this catalogue. We appreciate the fact that most of
the identifications of the Cuban insects were organized by
Dr Woody Horning a Curator at the Macleay Museum from
1982 to 1994. Dr Woody Horning identified much of the
insects himself and organized other American taxonomists
to identify material as well.
REFERENCES
Naumann I.D. and Steinbauer M.J. (2001). Egg parasitoids
of Australian Coreidae (Hemiptera). Australian
Journal of Entomology 40, 9-16.
Triplehorn, C.A. and Johnson NF (2005) “Borror and
DeLong’s Introduction to the Study of Insects.’
(Thomson Learning, Southbank, Victoria, Australia.)
Fletcher, J.J. (1920). The Society’s heritage from the
Macleays. Proceedings of the Linnean Society of New
South Wales 45, 567-635.
Macleay, W.S. (1838). ‘Illustrations of the Annulosa of
South Africa’. (Smith, Elder and Co., London).
Proc. Linn. Soc. N.S.W., 131, 2010
D9.
THE MACLEAY COLLECTION OF CUBAN BEETLES
CATALOGUE
Blattodea
FAMILY NUMBER
Sitio A oe al Ae eciaaed
cca =a | a Maa sei
tO
Wn
oo
Coleoptera
FAMILY
NUMBER
5
4
eae eee ee wha
iar er
0
r
ea one chal ads Ba
)
tO
4
— iw) —
ee ps i)
8
2
2
Carabidae
apie LT Es aa ETE
Cerambycidae | Amphidesmus | |
| Cerambycidae | Callichroma | |
Cerambycidae
Cerambycidae Eburia (orale nas thee
| Cerambycidae | Eburodacrys |
—
—=}h
Cerambycidae | Elateropsis
Cerambycidae Elateropsis fuliginosa Fabricius
Cerambycidae venusta Chevrolat
Cerambycidae
Cerambycidae | Elateropsis |
ee Le ate
Cerambycids [Leptosptis |
FCerambyeidae | Odontacea [|
|
Cerambycidae maculicornis Chevrolat 1862
Cerambycidae
30 Proc. Linn. Soc. N.S.W., 131, 2010
Cerambycidae Stenodontes
Cerambycidae
Chrysomelidae Cassida
Chrysomelidae Coptocycla
Chrysomelidae
Ciidae
Curculionidae Attelabus
Curculionidae Baridius madrimaculatus Boheman
Curculionidae Calandra
Curculionidae
Curculionidae Eurhinus
Curculionidae Exophthalmus
Curculionidae Exophthalmus
Curculionidae Exophthalmus
Curculionidae Exophthalmus
Curculionidae Exophthalmus
Curculionidae Hilipus
Curculionidae Hilipus
| Curculionidae Hilipus
Curculionidae Lachnopus
Curculionidae Lachnopus
Curculionidae Lachnopus
D. CROSS AND E. JEFFERYS
damicornis Linnaeus 1771 5
161
dorsopunctata Boheman 16
2
27
6
4
2
agaves 1
sericea Olivier 1807
Curculionidae 14
i
i)
pEmonee jee sl
haat ae
alee Od]
uous |
Jreyreissi Boheman 1836
guttatus Boheman 1843
rusticus Boheman 1836
curvipes Fabricius
N |W] tv
i
hispidus Gyllenhal
vittatus Gyllenhal
Curculionidae Lachnopus
Curculionidae Pachnéus
Curculionidae Pachnéus
Curculionidae Peltophorus
Curculionidae Polydacrys
Sle IVNIniIwlyel|wsy
—
azurescens Gyllenhal
litus Germar
NO] nn] dy
modestus Gyllenhal
Curculionidae Prepodes
Curculionidae Ptilopus
Curculionidae Rhina
Curculionidae Scyphophorus
Curculionidae Sphenophorus
Curculionidae Sphenophorus
Curculionidae Tetrabothynus
Curculionidae
Curculionidae
Curculionidae
Curculionidae
Dytiscidae Rhantus
Dytiscidae
Tetrabothynus
Tylomus
Xyleborus
Elateridae Pyrophorus
Histeridae
Lampyridae
Lycidae Calopteron
ite
| Mordetlidae |
3
9
Elateridae 2
1
5
;
spectabilis Dejean 14
vittatus Dejean
9
scrutator Olivier 4
1
atheniunus Schedl
sericeus Latreille
spectabilis Gyllenhal
calidus Fabricius 1792
phosphorescens
bicolor Linnaeus i
lon
Prog, Eimns Soc, NESW. 131, 2010
31
THE MACLEAY COLLECTION OF CUBAN BEETLES
Passalidae Passalus
Passalidae Passalus
Trox
| Trogossitidae |
UNIDENTIFED |
Diptera
ea ae
iain I a
[ONINENTREED|Ser
PUNIDENTIFIED | |_|
Hemiptera
GENUS | SPECIES | NUMBER
BBelasiomaticize) | ee aa | ae
Cees ae OC ae a ae ee 2
Cicadidae Cicada viridicincta Macleay 6
KE ae a a a ee
sexi ae ea a eee
Resa «Eo ee aE Se
Cascio s(t ee as Ee or)
SGennitize;etien "5 Eeiiseg)|_ 1Ab nent Daa
tiie Gon Caen
Membracidae
cS ee Er ee ea
Rec ae es
es aS a
Rio a
Gascon || cane ee eee
Pentatomidae
ee ee ee ee
Redividacle) )/|Uzioaa | es
eS ee eee
| Reduviidae |
UNIDENTIFIED
ened, ieee |
ENDEMED |e eoe | eee | omen
52. Proc. Linn. Soc. N.S.W., 131, 2010
D. CROSS AND E. JEFFERYS
FAMILY
Anthophoridae
Apidae
Apidae
ethylide os
@)
2
S)
S
AS)
g
ENUS SPECIES NUMBER
irate cere we
jimbriata Fabricius 1804
a)
<
3
@Q
i=)
lo}
as}
—
Q
5
i)
Braconidae
Chalcididae Brachymeria
Chalcididae Brachymeria
Chalcididae Brachymeria robusta Cresson 44
Chaleididae | Brachymeria | 47
Chaleididae 8
Chalcididae 3
debilis Say 1836 4
xanticles (Walker) 8
ARIE (CESSED) 19
Chalcididae Spilochalcis femorata (Fabricius) 8
Chalcididae Spilochalcis maniae (Riley) 2
| Chalcididae Spilochalcis
Chalcididae Spilochalcis
3
Chalcididae Spilochalcis 10
Chalcididae Spilochalcis
1
110
3
1
3
0
—
—
Chalcididae Spilochalcis
Chalcididae
Chrysididae Caenochrysis 1
Chrysididae Chrysis insularis Guérin 5
Chrysididae Chrysis 5
Chrysididae Chrysis 15
Chrysididae | Chrysis [| purpuriventris | |
Chrysididae | Chrysis
Chrysididae Holopyga ventralis Say 2)
Chrysididae 1
(pete)
Cynipidae a a a TE
ee |
aE
[Huser | se ew eee 2 es a TST
Geet a a ry
Formicidae
Formicidae 13
Formicidae Crematogaster 1
Formicidae Cyphomyrmex 1
Formicidae Odontomachus relictus 13
Camponotus
Proc. Linn. Soc. N.S.W., 131, 2010
THE MACLE
Megachilidae ii
Mutillidae
Platygastridae
Pompilidae Pepsis
Pompilidae
Scoliidae
AY COLLECTION OF CUBAN BEETLES
trifasciata Burmeister
trifasciata Burmeister
Scoliidae Elis
Sphecidae Monedula
Sphecidae Nysson
Sphecidae Nysson
insularis Dahl
albilabris
collaris
Nysson
Sphecidae
Nysson
hyalius
sericeus
Sphecidae
Sphecidae
Tiphiidae
Vespidae Ancistrocerus
| Vespidae | Eumenes
Vespidae Euodynerus
Pachodynerus [| | ST
9
cingulatus Cresson
7
2
50
Parancistrocerus | enyo (Lepeletier) 1841 27],
12
Vespidae Zeta pe
Vespidae Zethus 14
Lepidoptera
FAMILY GENUS SPECIES
Lycaenidae Cyclargus ammon (Lucas) 1857
Lycaenidae Eumaeus atala Poey 1832 ]
Lycaenidae Leptotes theonus (Lucas) 1857
Lycaenidae
Lycaenidae
Nymphalidae | Anaea
Nymphalidae Apatura pavonii Latreille 3
Nymphalidae Eunica 2
Nymphalidae Hypanartia | paullus Fabricius 1793 2)
Nymphalidae Megalura
Nymphalidae Metamorpha
IENpEERCRE[Eurcioddes
| Nymphalidae | Siderone |
Nymphalidae
Lo
4 Proc. Linn. Soc. N.S.W., 131, 2010
D. CROSS AND E. JEFFERYS
Papilio
Papilio
Papilionidae Papilio
Papilionidae Papilio
Papilionidae Papilio - ey
Db OuUVd
Pieridae
Sphingidae
UNIDENTIFIED
UNIDENTIFIED
Neuroptera
FAMILY
Myrmeleontidae
UNIDENTIFIED
_ Odonata
as 200
Orthoptera
iran
imam) |
NUMBER |
[psa
Phasmatodea
FAMILY GENUS | SPECIES | NUMBER
UNDENDIED | ot. | 20
eee
| Pulicidae | Ctenocephalides | felis (Bouché) 1836 | 8 |
[Tee PE a ee Te
Proc. Linn. Soc. N.S.W., 131, 2010
oe LAP tts ORGANS Hi o11,.4
ee i aha | om * —_————
ipoiny, ‘ae
r Diner @ ee | eer ‘ : psn
My \ La vr eet, bh yea 5
— age Saal a gtr Gem b+ IE |
| Np means ig
; SH, eS ee ee ie ati
¢ oy ; Chey Gaps) qn:
0 io at So
re oe
*Y {ahi A LerwmT}*
y [et & naan oy
|
e
‘ r
; _ ai i ® ' : ase | 4 nya +
oye Nic mae ae Seen? Be 2 a —~
- 7 a
* ul
7 - %
a vi — - ee
\
_ > _— ee
} : at
a! i
‘ie <
; t
; an ih, meer tr ale ay See
‘ Tey 4 is { er ,
; TSM (oT ase) RU
Tae ae 1 — rer © a ee
i T F i
Li i j
4 , hi gt —— - i nleermie f
Laie | ti £0
ai Thay : ‘ ! i
i x Py i i prey cP "pee on ay slat Car
5 ) i 1 =
. APES)
J i } i :
a 4,10 a)
¥ J
i
i gi i
< => > ia
’ oO )
r
oo |
es “
7
” -
: ,
V
bind 1 i
i i
' .
N
i i
; ke if Ti
nl r i)
» ¢ m
I
\ jaa uy 7
Fr) A amy Bes
1 te
a i Ve '
iis = Fu
ees
5 fhe ie Blea ‘haw
} "
\ ry ‘
i Al
a
P 7 fi i
a: i
]
a fa t
-
ay oe bs)
bok
aed: ee ete : Tr fom : D a a
Aan i ea cel ‘ P r “ ai pte a
rey Pik ha ie i ey ' ae PORES J} Pe anne
} : Pia —_ . } tr
— tg at
Description of a New Species of Inola Davies (Araneae:
Pisauridae), the Male of I. subtilis Davies and Notes on Their
Chromosomes
MartIN Tio! AND MARGARET HUMPHREY~*
'Faculty of Medicine, University of Sydney, NSW 2006, Australia
*Australian Museum, 6 College Street, NSW 2010, Australia
*Corresponding author (margaret.humphrey@yahoo.com.au)
Tio, M. and Humphrey, M. (2010). Description of a new species of Inola Davies (Araneae: Pisauridae),
the male of Z. subtilis Davies and notes on their chromosomes. Proceedings of the Linnean Society of
New South Wales 131, 37-42
A new pisaurid spider, Inola daviesae sp.n. is described from northern Queensland together with the first
description of the male of /. subtilis. The meiotic chromosomes of both species are discussed.
Manuscript received 5 March 2010, accepted for publication 21 April 2010.
KEYWORDS: chromosomes, /no/a, Pisaurid, Queensland, rainforest, spider.
INTRODUCTION
Australian pisaurid spiders are generally not
web builders, except for members of J/no/a Davies,
1982 and Dendolycosa Koch, 1876. The genus Jnola
includes three species from northeastern Queensland
( Davies, 1982). Like Davies’ Inola species, Inola
daviesae sp.n. described here is a delicate, medium-
sized spider associated with tropical rainforest. As
with other members of the genus, this spider runs on
the upper surface of its horizontal sheet web. These
webs project from the trunks of rainforest trees or
embankments. A short silken funnel extends from the
sheet web to a retreat in a tree trunk or embankment.
The females, like those of other pisaurids, grasp their
egg sacs in their chelicerae when disturbed and carry
them into their retreat (Davies, 1982).
Abbreviations: CL cephalothorax length; CW
cephalothorax width; AL abdomen length; AW
abdomen width; MOQ median ocular quadrangle;
AM Australian Museum; QM Queensland Museum.
MATERIALS AND METHODS
Morphology
Measurements were made with an ocular
micrometer and converted to millimetres.
Measurements are for a single specimen with a range
of variation if significant. Spines have been recorded
as number per surface for each segment, as they were
often staggered.
Chromosomes
Live penultimate male spiders were anaesthetized
with CO,. The testes were dissected out and sections
were spread, fixed and stained after the method of
Rowell (1991). These preparations were viewed and
photographed using a light microscope. Counts and
other observations were noted from photographs of
many (>50), suitable meiotic cells in metaphase and
chromosome numbers for species determined by the
mode.
NEW SPECIES OF JNOLA
SYSTEMATICS
Genus /nola Davies, 1982
Inola Davies, 1982: 479
Type species: /nola amicabilis Davies, 1982, by
original designation (page 480).
Inola daviesae n.sp.
Figs (1-4, 7-11)
Types
Holotype: male, Leo Creek, Macllwraith Ranges,
North Qld. [13°32’S 143°29’E], July, 1995, M.
Humphrey, M. Moulds, KS58316 (AM). Paratypes:
1 female, same data as holotype, KS58322 (AM); 1
male, 5 females, Qld. MaclIlwraith Ranges, Leo Creek
[13°32’S 143°29°E], 20 Jul 1995, M. Humphrey, M.
Moulds, F. MacKillop, KS43933 (AM); 1 male, 1
female, data as for holotype, QMS 83903 (QM).
Other material examined
Eleven juveniles, same data as holotype,
KS58315 (AM).
Distribution
Rainforest, MaclIlwraith Range, north-eastern
Queensland at an altitude of approximately 500m.
Diagnosis
Males can be distinguished from other members
of the genus by the distinctive spannerhead-shaped
distal portion of the median apophysis of the male
palp (Fig. 8). The female scape is narrow while that of
I. cracentis is broad and that of /. subtilis is triangular,
pointed posteriorly and broad anteriorly.
Description of male
Measurements of holotype: CL 4.3, CW 2.7, AL
5.9, AW 1.7. Eye group: anterior width 1.1; posterior
width 1.1; length 0.6; MOQ: anterior width 0.4;
posterior width 0.5; length 0.5. Maxilla: length 1.3;
width 0.8; Sternum: length 1.9; width 1.9; Colulus:
length 0.2; width 0.3. Leg lengths:
Palp i 2 3 4
Femur 4.5 11.9 11.4 7.6 eS
Patella 1.6 1.9 2.0 1.4 eS
Tibia 1.9 ital ES 8.9 10.3
Metatarsus
—_ Sal 14.6 9.8 15.4
Tarsus 4.6 4.9 4.6 3:5 4.9
Total 12.6 44.9 43.9 a2 43.4
38
Spine notation: Palp: femur, d3p1; patella, dSpI1r1;
tibia, d2r2; tarsus, 0. Leg I: femur, d4p5; patella, d1;
tibia, d3p3r2v3; metatarsus, p4r4v1, whorl of four
small spines distally ; tarsus, 0. Leg II: femur, d2p5r5;
patella, dl; tibia, d2p3r4v3; metatarsus, p3r4, whorl
of four small spines distally; tarsus, 0. Leg III: femur,
d2p4r5; patella, dl; tibia, p2r4v3; metatarsus, p4r4,
whorl of four small spines distally; tarsus, 0. Leg IV:
femur, d3p4r2; patella,d1; tibia p3r3v2; metatarsus,
d4p4; tarsus,0. Note: four distal spines on end of each
metatarsus.
Eye diameters roughly equal. Cephalothorax
patterned (Fig.l). Abdomen with centrai pale stripe
to almost half the length of abdomen. Pair of pale
latero-dorsal stripes, running three quarters of the
abdomen. Two or three pairs of prominent pale spots
between the central and the latero-dorsal stripes. Legs
banded.
Palp (Figs.7, 8). Digitiform portion half the length
of the palpal tarsus. Median apophysis large and partly
membraneous, partly sclerotised. Distal sclerotised
portion bifid (spanner-like). Embolus slender and
curved. Conductor behind median apophysis with a
fold distally.
Description of female
Measurements of KS58322: CL 3.9, CW 3.4, AL
6.9, AW 4.7. Eye group: anterior width 1.5; posterior
width 1.6; length 1.0; MOQ: anterior width 0.7;
posterior width 0.8; length 0.7. Maxilla: length 1.6;
width 1.0. Sternum: length 2.6; width 2.1; Colulus:
length 0.2; width 0.3. Leg lengths:
Pal 1 2 3 4
Femur 2.8 8.6 9.4 Tes) 9.4
Patella 1.0 DD, 1.9 1.6 1.6
Tibia 1.4 8.8 93 6.1 8.0
Metatarsus
— 10.4 8.6 7.9 12.5
Tarsus 3.1 3.0 BS 3.1 43
Total 8.3 33.0 B2a/ 26.2 35.8
Spine notation: Palp: femur, d1p1; patella, d1p1;
tibia d2p2rl, tarsus, p2. Leg I: femur, d2r2; patella,
dl; tibia, dlr2vl; metatarsus, d3r4v2, whorl of four
small spines distally, tarsus, 0. Leg II: femur, d2p5r5;
patella, dl; tibia, dlp2r2vl; metatarsus, d3p2r4v2,
whorl of four small spines distally; tarsus, 0. Leg III:
femur, d4p2v1; patella, dilrl; tibia, 0; metatarsus,
dlp2r2v2, whorl of four small spines distally;
tarsus, 0. Leg IV: femur, d4r5; patella dirl; tibia,
0; Metatarsus, d2p3rlv2, whorl of four small spines
distally; tarsus, 0. Note: four distal spines on end of
metatarsus (every leg).
Proc. Linn. Soc. N.S.W., 131, 2010
M. TIO AND M. HUMPHREY
Figures 1-7. 1, Inola daviesae sp. n. male carapace, dorsal, (holotype). 2, Imola daviesae sp.n. male cepha-
lothorax, lateral, (holotype). 3, Inola daviesae sp. n. epigynum, external, (KS58322). 4, Inola daviesae sp.
n. epigynum, internal, ventral. 5, Inola subtilis, male palp, ventral, (KS58321). 6, Inola subtilis, expanded
male palp, retrolateral, (KS58320).
Proc. Linn. Soc. N.S.W., 131, 2010 39
NEW SPECIES OF INOLA
30 jim 11 va 2opm 12
Figures 7-12. 7, male palp of Jnola daviesae sp.n. 8, median apophysis (ma), embolus (e) and conductor
(c) of Inola daviesae sp. n. 9, Inola daviesae sp.n. female on sheet web. 10, Inola daviesae sp.n., prophase
male meiotic chromosomes showing two dense sex chromosomes (arrowed). 11, Inola daviesae sp.n., male
meiotic cell showing 14 pairs of chromosomes. 12, Inola subtilis, male prophase meiosis showing two
densely stained sex chromosomes (arrowed).
40 Proc. Linn. Soc. N.S.W., 131, 2010
M. TIO AND M. HUMPHREY
Epigynum (Figs 3, 4). Scape a narrow bar.
Insemination ducts arise near hind edge of the
epigastrum and travel forward. Large stalked
spermathacae. Insemination duct enters near the
base of the posterior spermathacae (fertilisation duct
leaves below this junction).
Chromosomes
For males of /. daviesae sp. n., 2N = 28 (Fig.
11), including two subequal, darkly staining sex
chromosomes. Most of the 13 pairs of autosomes in
Inola daviesae sp. n. appear to be telocentric. The
two sex chromosomes are easily distinguished in
prophase of meiosis (Fig. 10). They migrate from
the equator of the spindle in metaphase as a pair and
earlier than the autosomes. Such sex chromosomes
and their behaviour have been observed in other
spiders by Rowell (1991). According to a survey
of spider chromosome studies, (Rowell, personal
comm.), female spiders have double the number of
_ sex chromosomes to those of the male. Presuming
this species follows the same sex determination
mechanism, males of /nola daviesae n. sp. would be
XX and females XXXX, giving females 2N = 32.
Etymology.
Named for Valerie Todd Davies who described
the genus.
Inola subtilis Davies, 1982
(Figs 5, 6)
Material examined
1 male, Goldsborough S. F., Qld., July, 1995,
M. Humphrey, KS58321 (AM); 1 male, data as for
KS58321, QMS83902 (QM); 3 males, data as for
KS58321, KS58320 (AM); 1 male, Palm Cove,
FNQ, J.Olive, 6 Sept 1995, sheet web on fallen log,
KS044108 (AM); Goldsborough Valley SF, rainforest
strangler fig, 27 Jul 1995, M. Humphrey, KS043900
(AM).
Distribution
Material from Davies’ description of the
species indicates a distribution on the western edge
of suburban Cairns. The material examined above
extends this distribution from Palm Cove (north of
Cairns) to the Goldsborough Valley in the south.
Diagnosis for male
Unlike the other three members of the genus, the
sclerotised distal portion of the male palpal median
apophysis forms two, fused, parallel, curved processes
Bro. mn, Soc. N:SAW.A3 15 20110
(Fig. 5). Proximally is a long, narrow sclerotised
spur pointing ventrally, at right angles to the palp.
Conductor sclerotised, retrolateral, behind the large
median apophysis and bearing a spine distally.
Description of male
Measurements of KS58321: CL 3.5, CW 2.8, AL
4.4, AW 1.44; Eye group: anterior width 0.8; posterior
width 1.2; length 0.8; MOQ: anterior width 0.5,
posterior width 0.6, length 0.5. Maxilla: length 1.0;
width 0.5. Sternum: length 1.8, width 1.7. Colulus:
length 0.3, width 0.5. Leg lengths:
Palp 1 2 3 4
Femur 2.0 10.0 9.3) es) 9.4
Patella 0.6 1.6 1.6 1.5 1.5
Tibia 0.8 10.3 9.4 6.9 8.9
Metatarsus
— 12.6 12.1 9.1 13.0
Tarsus 1.6 3.9 3.6 2.5 353)
Total 5.0 38.4 36.0 DS 36.1
Spine notations: Palp: femur, d2p1; patella,
dipIirl; tibia, d2p2; tarsus, 0. Leg I: femur, d2p8r3;
patella, dl; tibia, d3p2r2v4; metatarsus, d3p2r5v2;
tarsus, 0. Leg I: femur, p5; patella, d5p6r3; tibia,
d2p2r2v3; metatarsus, dlp3r3v1; tarsus, 0. Leg III:
femur, d2p5r5; patella, d1; tibia, d2p3r3v3; metatarsus,
d2p2r2v2; tarsus, 0. Leg IV: femur, d2p5r2; patella,
d2p4r2v3; tibia, dlplr3vl; metatarsus, d2p2r2;
tarsus, 0.
Abdomen long and narrow. Abdominal pattern
with pale centre stripe and a pair of pale latero-dorsal
stripes. Pairs of prominent pale spots as in 1. daviesae
but spots continue in line and merge to form a pair of
additional stripes. Legs banded.
Male palp (see diagnosis): Length of digitiform
portion almost half of palpal tarsus. Embolus curved,
slender, lying between median apophysis and
conductor.
Chromosomes
Because of poor spreading, the number of
chromosomes of J. subtilis could only be estimated.
However, it is between 26 and 32 and most of the
chromosomes are telocentric. There are two sex
chromosomes (Fig. 12) and like those of 1. daviesae
sp. n., they are darkly staining and migrate from the
equator of the spindle earlier than the autosomes.
41
NEW SPECIES OF INOLA
ACKNOWLEDGMENTS
We are grateful to the following staff and departments
of the University of Sydney; Assoc. Prof. L.W. Burgess,
Dean, Faculty of Agriculture, for provision of a mentorship
to the senior author; Assoc. Prof. H. A. Rose, Department of
Crop Sciences and the Electron Microscope Unit for the use
of their facilities. Our thanks also to Dr Valerie Todd Davies,
Queensland Museum, for specimen identifications and to
Dr M. Gray and Dr M.S. Moulds, Australian Museum, for
valuable advice and assistance.
REFERENCES
Davies, V. T. (1982). /nola nov. gen., a web-building
pisaurid (Araneae: Pisauridae) from northern
Australia with descriptions of three species. Memoirs
of the Queensland Museum 20(3): 479-487.
Rowell, D.M. (1991). Chromosomal fusion in De/ena
cancerides (Araneae: Sparassidae). I. Chromosome
pairing and X-chromosome segregation. Genome 34:
561-573.
42 Proc. Linn. Soc. N.S.W., 131, 2010
A Late Ordovician Conodont Fauna from the Lower Limestone
Member of the Benjamin Limestone in Central Tasmania, and
Revision of Tasmanognathus careyi Burrett, 1979
Y.Y. ZHEN!, C.F. BurretT’, I.G. PERCIVAL? AND B.Y. Lin*
‘Australian Museum, 6 College Street, Sydney, N.S.W. 2010, Australia (yongyi.zhen@austmus.gov.au);
*School of Earth Sciences, University of Tasmania, GPO Box 79, Hobart, Tasmania 7001, Australia
(cliveburrett@gmail.com);
*Geological Survey of New South Wales, 947-953 Londonderry Road, Londonderry, N.S.W. 2753, Australia
(ian.percival@industry.nsw.gov.au);
“Institute of Geology, Chinese Academy of Geological Sciences, Beijing, China, 100037.
Zhen, Y.Y., Burrett, C.F., Percival, I.G. and Lin, B.Y. (2010). A Late Ordovician conodont fauna from
the Lower Limestone Member of the Benjamin Limestone in central Tasmania, and revision of
Tasmanognathus careyi Burrett, 1979. Proceedings of the Linnean Society of New South Wales 131,
43-72.
Ten conodont species, including Aphelognathus? sp., Belodina compressa, Chirognathus tricostatus
sp. nov., Drepanodus sp., gen. et sp. indet., Panderodus gracilis, Protopanderodus? nogamii, Phragmodus
undatus, Tasmanognathus careyi and T. sp. cf. T. careyi are documented from the Lower Limestone
Member of the Benjamin Limestone, Gordon Group, exposed in the Florentine Valley and Everlasting Hills
region of central Tasmania. For the first time since its establishment three decades ago, the type species of
Tasmanognathus, T. careyi, is revised with recognition of a septimembrate apparatus including makellate M,
alate Sa, digyrate Sb, bipennate Sc, tertiopedate Sd, carminate Pa, and Pb (angulate Pb! and pastinate Pb2)
elements. Co-occurrence of Phragmodus undatus and Belodina compressa in the fauna indicates a latest
Sandbian to earliest Katian (Phragmodus undatus conodont Zone) age for the Lower Limestone Member
of the Benjamin Limestone. All species previously attributed to Zasmanognathus are briefly reviewed,
and the distribution of the genus is shown to be more widespread than hitherto recognised (in New South
Wales, North China, Tarim Basin, South Korea and northeast Russia), with a probable occurrence in North
American Midcontinental faunas.
Manuscript received 16 September 2009, accepted for publication 26 May 2010.
KEYWORDS: Benjamin Limestone, biogeography, biostratigraphy, conodonts, Late Ordovician,
Tasmania, Zasmanognathus.
INTRODUCTION
Ordovician conodont faunas of Tasmania are
relatively poorly known in comparison to those from
the mainland of Eastern Australia. Only three papers
— Burrett (1979), Burrett et al. (1983) and Cantrill
and Burrett (2004) — have dealt systematically with
a small number of species. The present contribution,
which describes the comparatively diverse fauna
from the lower part of the Benjamin Limestone, is the
first part of a revision of all known conodonts from
Tasmania. This project aims to provide a firm basis
for conodont-based correlations of the carbonate-
dominated Gordon Group with limestones along the
Delamerian continental margin in New South Wales,
with strata in offshore island arc settings in central
N.S.W. (Macquarie Arc), and with isolated limestone
pods in the New England Orogen in northeastern
N.S.W. and central Queensland.
Given the rarity of graptolites in the
predominantly shallow-water platformal succession
forming the Delamerian margin succession, and
the sparsely documented occurrences of conodonts,
LATE ORDOVICIAN CONODONTS FROM TASMANIA
biostratigraphical zonation in Ordovician rocks
of Tasmania is currently largely reliant on shelly
macrofossils. Banks and Burrett (1980) established
a series of twenty successive faunas (designated
OT assemblages 1-20), one of which (OT 12) was
defined by the occurrence of several conodont species
including Tasmanognathus Chirognathus
monodactyla, Erismodus gracilis and Plectodina
aculeata in the basal Benjamin Limestone. This
fauna (based at the time on unpublished studies by
Burrett, with no species illustrated or described
in the 1980 paper) is revised here. Our study has
not identified the last two named species, and has
recognised a new species of Chirognathus in place
of C. monodactyla. Burrett (in Webby et al. 1981,
p.12) summarised the occurrences of conodonts in
the Tasmanian Ordovician succession. He noted the
first appearance of the biostratigraphically important
species Phragmodus undatus in strata immediately
above the Lords Siltstone Member in the middle of the
Benjamin Limestone; however, our reassessment of
the fauna has identified the presence of this species in
the underlying lower part of the Benjamin Limestone.
Laurie (1991) defined an alternate series of 20 faunal
assemblages based on Tasmanian brachiopods,
ranging in age from Early Ordovician (Tremadocian)
to earliest Silurian. Where possible, these brachiopod
faunas were tied in to conodont occurrences, mainly
derived from Burrett’s (1978) unpublished thesis
studies.
A biogeographically significant component of
the Tasmanian conodont fauna is 7asmanognathus
Burrett, 1979, which was first identified from the
Lower Limestone Member of the Benjamin Limestone
exposed in the Florentine Valley and Everlasting Hills
region of central Tasmania (Fig. 1). This genus has
subsequently been widely recognized as occurring
in rocks of early Late Ordovician (Sandbian) age in
eastern Australia and China. Low yields (averaging
two specimens per kg) of conodonts from the Gordon
Group carbonates collected and processed by Burrett
(1978) resulted in Jasmanognathus being imperfectly
defined. Thirty years after its initial documentation,
revision of the type species, 7. careyi Burrett, 1979
has become urgently needed in order to better
understand its multielement apparatus, phylogenetic
relationship and precise stratigraphic range in the
type area. The purpose of this paper is to describe
the conodont fauna from the middle part of the
Gordon Group in the Settlement Road section of the
Florentine Valley area, equivalent to the level yielding
Tasmanognathus, based on five recently collected
bulk samples of limestone totalling 49.5 kg that on
dissolution in acetic acid have yielded an average
careyl,
44
of six elements per kg. These additional collections
are supplemented by re-examination of Burrett’s
original material including types and topotypes of
T. careyi, and for the first time all the accompanying
conodont fauna is documented by description and/or
illustration, including Aphelognathus? sp., Belodina
compressa (Branson and Mehl, 1933), Chirognathus
tricostatus sp. nov., Drepanodus sp., Panderodus
gracilis (Branson and Mehl, 1933), Protopanderodus?
nogamii (Lee, 1975), Phragmodus undatus Branson
and Mehl, 1933, and gen. et sp. indet.
REGIONAL GEOLOGIC AND
BIOSTRATIGRAPHIC SETTING
Platform sedimentary rocks of the Early
Palaeozoic Wurawina Supergroup, that are
widespread in the western half of Tasmania, consist
of the Late Cambrian — Early Ordovician Denison
Group (mainly siliciclastics), conformably overlain
by the Gordon Group (predominantly carbonates of
Early to Late Ordovician age), in turn conformably
or disconformably overlain by the Hirnantian (latest
Ordovician) to mid-Devonian Eldon Group, which
consists mainly of siliciclastics (Burrett et al. 1984;
Laurie 1991). The Gordon Group attains a thickness
of 2100m of carbonates and minor siltstones in
its redefined type section in the Florentine Valley
where it is divided into three limestone formations.
The uppermost of these, the Benjamin Limestone,
is divided into two limestone members (Upper and
Lower) separated by a thin but regionally extensive,
macrofossiliferous siltstone member (Lords Siltstone
Member). The Benjamin Limestone predominantly
consists of interbedded microcrystalline peritidal
dolomitic micrite, dolostone and calcarenite with
a maximum thickness of about 1200m. Some 400
conodont samples were initially collected over a Sm
interval by Burrett (1978) from the various localities
of the Gordon Group, but many of these samples were
barren or had a very low yield, due to the peritidal
to shallow subtidal depositional setting and high rate
of sedimentation in the tropical shelf environments.
Continuous efforts in the last 30 years by post-
graduate students and academic staff of the University
of Tasmania have accumulated significant amounts
of conodont material for the age determination
and biostratigraphic analysis of the Gordon Group
(Burrett 1979; Burrett et al. 1983, 1984; Cantrill and
Burrett 2004).
Carbonates that are coeval with the Lower
Limestone Member of the Benjamin Limestone occur
in many sections in northern, western and southern
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
eEverlasting
Hills
Florentine e
Valley § p
np
aa
* Sample locations EG Gordon Group ee
Lanes
4Peak 0
Gells 4
Lookout
MT FIELD
a Mt Field
West %
ides %Z.
nun ee
AS?” NATIONAL % PARK
Florentine
i: > an
River \ NS
loidg
~~ rt
Sandstone Sandstones
(Early? Ordovician) Mabon (late Ordovician to Silurian?)
Figure 1. Maps showing the studied areas in central Tasmania and sample locations. A, Map of Tasma-
nia showing the locations of Florentine Valley and Everlasting Hills (from Burrett 1978, 1979); B, Map
showing the Florentine Valley area and sample location of the Nine Road Section (modified from Laurie
1991); C, Map showing the Settlement Road Section of the Florentine Valley and sample locations (modi-
fied from Laurie 1991); D, Map showing Everlasting Hills area and sample location (from Burrett 1978).
Tasmania, but the Zasmanognathus careyi fauna has
only been definitely found in the Florentine Valley
and in the Everlasting Hills. The Florentine Valley
sections (Figs | and 2) are found in the eastern side of
a mid-Devonian synclinorial structure. This area was
first mapped geologically by Corbett and Banks (1974)
and because of its completeness, has subsequently
been the focus of numerous palaeontological and
sedimentological studies. However, active timber
logging in this area has meant that some sections are
now inaccessible, having been replanted with dense,
almost impenetrable, forest.
The Everlasting Hills section (Fig.1D) was
discovered in remote and moderately dense to thick
Proc. Linn. Soc. N.S.W., 131, 2010
vegetation and mapped by Ian McKendrick and Clive
Burrett in 1975 (Fig.1D). This doline and cave-rich area
has since been included in the South West Tasmania
World Heritage wilderness area, and has undergone
extensive regrowth so that it is now extremely difficult
to access. The palaeotropical limestones in the
Everlasting Hills are identical to those in the Lower
Limestone Member of the Benjamin Limestone in the
Florentine Valley, and consist of 3-6m thick Punctuated
Aggradational Cycles (Goodwin and Anderson 1985)
of mainly dolomitised, intertidal micrites with tidal
channels and top beds containing a lower intertidal
to high subtidal macrofauna. Somewhat deeper water,
coeval carbonates (the Ugbrook Formation) occur in
45
ORDOVICIAN CONODONTS FROM TASMANIA
»
4
LATE
FLORENTINE VALLEY
SETTLEMENT ROAD
FLORENTINE VALLEY
EVERLASTING HILLS
NINE ROAD
tAaieo 1 39 ‘ds snyjeuBouewse; en &
iAaieo snyjeuBouewse) a &
snjepun snpowBeiyd ee
timeBou gsnposapuedojosg eeep
siioei6 snposepueg eS &
yepui “ds ja -uag @
‘ds snpouedaiq ; e<
snjejsoou) snyjeuBouiyD ORR
essaidwoo eulpojag ry
‘ds gsnyjeubojoydy e
LORDS
SILTSTONE
MEMBER
CASHIONS
CREEK
LIMESTONE
tAaueo jy (yo ‘ds snyjeuBouewse; e
iAeseo snyjeuBouewsey ee
snjepun snpowBeiy4 e—e
sijoeiB snposepueg eo—_e
‘ds snpouedaig e
snjejsoouy snyjeuBosy9 ee
essaidwos eulpojag e
thaieo | jo “ds snyjeuBouewse; e_eo0
ifaieo snyjeuBouewses e- ee
snjepun snpowBbeiyd &
‘ds snpouedaiq ee =
snjejsoou} snyjeuBouiyD e S
‘ds gsnyjeubojaydy @
uw
a6 9 ¥
rn YaaWAW YaddN = one Zo
ra a Baad ANOLSAWI1 SYaaWyvy Be
iz YagW3W YaMO1 OSw 20
g°=
so JNOLS3WI1 NINVPN3a = 5°
dnouS NOGYOS
Figure 2. Three stratigraphic sections showing the sample horizons and ranges of the conodont species
in the Lower Limestone Member of the Benjamin Limestone, Gordon Group, in central Tasmania.
Proc. Linn. Soc. N.S.W., 131, 2010
46
YY ZHEN, CE BURREDIE EGSPERCIVAL AND BLY: LIN
northern and western Tasmania (Burrett et al. 1989)
but these lack Tasmanognathus. This suggests that
Tasmanognathus was mainly restricted to peritidal
tropical environments in the Late Ordovician.
The TYasmanognathus fauna is associated with a
strongly endemic macrofauna in the lower and middle
parts of the Lower Limestone Member, Benjamin
Limestone, including the brachiopods Lepidomena
Laurie, 1991, Yasmanorthis Laurie, 1991 and the
nautiloids Gorbyoceras settlementense Stait and
Flower, 1985, Paramadiganella Stait, 1984 and
Tasmanoceras zeehanense YVeichert and Glenister,
1952 (Laurie 1991; Stait 1988). Tasmanognathus
careyi is found in two of the twenty Ordovician
brachiopod assemblages (or biozones) recognised
by Laurie (1991); the Zasmanorthis calveri and the
younger Zasmanorthis costata assemblages.
AGE AND CORRELATION OF THE FAUNA
In the conodont fauna associated with
Tasmanognathus careyi from the Lower Limestone
Member of the Benjamin Limestone in central
Tasmania, occurrence of Phragmodus undatus and
Belodina compressa is crucial for age determination
and regional correlation, as both species are
cosmopolitan and age diagnostic. The former had a
relatively long stratigraphic range, extending from the
base of the Ph. undatus Zone (in the upper Sandbian)
to the top of the Katian, and the latter first occurs
at the base of the B. compressa Zone and extends
to the base of the B. confluens Zone (Sweet 1988).
Co-occurrence of these two species and absence of
any diagnostic species of either the B. confluens or
P. tenuis zones indicates a latest Sandbian to earliest
Katian age (Phragmodus undatus Zone) for this
Tasmanian fauna.
Chirognathus is also morphologically
distinctive with the two previously-reported species
(Chirognathus duodactylus Branson and Mehl,
1933 and Chirognathus cliefdenensis Zhen and
Webby, 1995) restricted to the upper Sandbian-
Katian interval (Sweet 1982; Zhen & Webby 1995).
The new species from Tasmania described herein is
morphologically similar to the type species of the
genus, C. duodactylus Branson and Mehl, 1933.
This species with a well-known multi-element
apparatus is widely distributed in Sandbian strata of
the North American Mid-continent ranging from the
Pygodus anserinus Zone to the Phragmodus undatus
Zone (Sweet in Ziegler 1991). The second species,
Chirognathus cliefdenensis Zhen and Webby, 1995,
occurs in a stratigraphically slightly younger interval
Proc. Linn. Soc. N.S.W., 131, 2010
in central New South Wales, where it is recorded
from the upper Fossil Hill Limestone to the lower
Vandon Limestone (early Katian) of the Cliefden
Caves Limestone Subgroup (Zhen and Webby 1995),
from the Downderry Limestone Member (late Katian)
of the Ballingoole Limestone of the Bowan Park
Limestone Subgroup (Zhen et al. 1999), and from
allochthonous limestones of Katian age emplaced in
the Silurian Barnby Hills Shale (Zhen et al. 2003a).
The Lower Limestone Member of the Benjamin
Limestone exposed in the Everlasting Hills and
Florentine Valley areas in central Tasmania is the
type stratum of Yasmanognathus careyi Burrett,
1979. Since the initial documentation of this species,
at least ten additional species from lower Sandbian
to upper Katian strata predominantly of North China
and eastern Australia have been accommodated in
Tasmanognathus (see Systematic section for further
discussion). The origin and phylogenetic relationships
of Yasmanognathus remain uncertain as most of
these species were poorly documented and need to
be revised. Reassessment of 7. careyi herein suggests
that Zasmanognathus may be closely related to so-
called “Ordovician ozarkodinides” (Sweet 1988,
p. 91-92), an informal group including forms like
“Plectodina’, Aphelognathus and Yaoxianognathus.
Based on similarities of their general morphology and
apparatus construction, Zasmanognathus, as a sister
group, seems closely related to Yaoxianognathus.
Tasmanognathus is potentially the direct ancestor
of the latter, which was mainly restricted to eastern
Gondwana and peri-Gondwanan terranes during
the Late Ordovician (Katian). Strong biogeographic
similarities (including Zasmanognathus) between the
North China Terrane (or block) and eastern Australia
were part of the evidence used by Burrett et al. (1990)
to suggest that these blocks were contiguous or
closely proximal during the Ordovician.
Tasmanognathus was widely reported from
the Sandbian in North China with recognition of
three biozones based on the inferred lineage of
Tasmanognathus species (An and Zheng 1990;
Lin and Qiu 1990), from the oldest 7: sishuiensis
Zhang in An et al., 1983 from the upper Fengfeng
Formation (lower Sandbian), to 7. shichuanheensis
An in An et al., 1985 from the middle-lower part of
the Yaoxian Formation (upper Sandbian), and then to
the youngest 7: multidentatus An in An and Zheng,
1990 (the latter is a nomem nudum, equivalent to T.
borealis An in An et al. 1985; see Systematic Section
for further discussion) from the upper part of the
Yaoxian Formation (upper Sandbian-lower Katian).
An and Zheng (1990, p. 95, text-fig. 9) illustrated
the morphological changes from T. sishuiensis with a
AT
LATE ORDOVICIAN CONODONTS FROM TASMANIA
robust cusp and small, widely spaced denticles on the
processes of the S elements, to 7) mu/tidentatus with
a small, indistinct cusp in the Pa element and closer
the S elements. Importantly, similar morphological
changes have also been observed between the two
species of Zasmanognathus recognized in the Lower
Limestone Member of the Benjamin Limestone in
central Tasmania. A species described herein as 7.
sp. cf. 7. careyi that bears a prominent cusp in the
Pa element and small, widely spaced denticles
on the processes of the S and Pb elements is more
comparable with 7: shichuanheensis from the middle-
lower part of the Yaoxian Formation, whereas 7:
careyi with a small or indistinct cusp in the Pa element
and long, closely spaced denticles on the processes of
the S elements is closer to 7: multidentatus from the
upper part of the Yaoxian Formation. 7. careyi was
also reported from the middle part of the Yaoxian
Formation in association with 7. shichuanheensis and
Belodina compressa in Bed 3, about 44 m below the
first occurrence of 7. multidentatus (An and Zheng
1990, p. 86-87), although An’s identification cannot
be confirmed without re-examination of the original
material (An et al. 1985) and further investigations.
Occurrence of Taoqupognathus blandus at the
top of the Yaoxian Formation in the Taoqupo Section
of Yaoxian County (formerly Yaoxian; An and
Zheng 1990) suggests that the Yaoxian Formation
may well extend to the lower Katian. Therefore, the
morphological characters shown by the two species of
Tasmanognathus from the Lower Limestone Member
of the Benjamin Limestone support a correlation
between this limestone unit in central Tasmania, and
the middle part of the Yaoxian Formation in North
China (with the possible occurrence of T. careyi),
which An and Zheng (1990, p. 92, table 2) correlated
with the C. wilsoni graptolite Zone (late Sandbian).
An and Zheng (1990, p.115) suggested that the
Llandoverian conodonts illustrated by Lee (1982)
from the Hoedongri Formation in the Taebaeksan
Basin, Kangweon-Do of South Korea were
comparable with the Tasmanognathus sishuiensis
assemblage from the upper Fengfeng Formation of
North China. In fact, in their revision of Lee’s original
identifications (An and Zheng 1990, table 5, pp. 118-
119), they believed what Lee (1982) illustrated as
Pterospathodus celloni (Walliser) should belong to
Tasmanognathus sishuiensis, and considered that the
Hoedongri Formation should be correlated with the
Baduo Formation or the upper part of the Fengfeng
Formation (Sandbian) of North China.
MATERIAL AND SAMPLING LOCALITIES
The current study is based on 683 identifiable
specimens from 10 samples (See Table 1). Of these,
Table 1. Distribution of conodont species in the samples studied.
N
2 =
species
Aphelognathus? sp.
Belodina compressa 8)
Chirognathus tricostatus sp. nov. 36 6
Drepanodus sp. 20
Gen. et sp. indet.
Panderodus gracilis All vep2
Protopanderodus? nogamii
Phragmodus undatus Pas Seay)
Tasmanognathus careyi 156
Tasmanognathus sp. cf. T. careyi 3
Total 290 15
48
Sera ate Shak SOs
sy OF he ha
2 2 4
1 10
12 Digi leet ls oad 71
Pare ee sl 2, D5
3 3
4 3 2, 2
4 44 3 18 69
6 1 Sib Sil 116
209 265 26 Wian23 O5y— lilee2977
is} 4 if 14 2 34
2" 48" 26" 23 79936 120) 2) 683
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
378 specimens are Burrett’s (1979) original material
including types of Zasmanognathus careyi recovered
from five samples collected from the Florentine Valley
and Everlasting Hills sections (see Burrett 1979, p.
32, fig. 1 for sample locations and their stratigraphic
horizons within the Lower Member of the Benjamin
Limestone). Samples LLMB, C137 and C98 were
collected from the Lower Limestone Member of the
Benjamin Limestone exposed along the Nine Road
(Fig. 1B). The Lower Limestone Member of the
Benjamin Limestone is exposed as a 50m thick section
(at Grid Ref. DP202157; 42°16.4’S, 146°2.65’E)
to the north side of the Everlasting Hills (Fig. 1D).
Two samples (JRC 2 and JRB) from this location
produced relatively abundant conodonts (Table 1).
The remaining 305 specimens were recovered from
five large spot samples — YYF 1 (13 kg), YYF2 (8 kg),
YYF3 (10 kg), YYF4 (7.5 kg), and YYFS5 (11 kg) —
collected from the lower part of the Lower Limestone
Member of the Benjamin Limestone in the Settlement
Road section of the Florentine Valley area (Figs 1C,
Ds
SYSTEMATIC PALAEONTOLOGY
All photographic illustrations shown in Figures
3 to 17 are SEM photomicrographs of conodonts
captured digitally (numbers with the prefix IY
are the file names of the digital images). Figured
specimens bearing the prefix AM F. are deposited in
the type collections of the Palaeontology Section at
the Australian Museum in Sydney. All the syntypes
except one (UTG96863 not located; figured by
Burrett 1979, pl. 1, figs 17-18) and most of the other
specimens of Zasmanognathus careyi illustrated by
Burrett (1979) were relocated and made available for
the current study. They have been now transferred to
the Australian Museum collection, and a new AM F.
registration number has been allocated to each of the
specimens illustrated in this contribution.
The following species are documented herein
only by illustration as they are either rare in the
collection or have been adequately described
elsewhere in the literature: Aphelognathus? sp. (Fig.
3J-K), Drepanodus sp. (Fig. 3C-F), gen. et sp. indet.
(Fig. 3G-I), and Panderodus gracilis (Branson and
Mehl, 1933) (Fig. 6A-I). Authorship of the new
species Chirognathus tricostatus is attributable solely
to Zhen. Taxa documented herein are alphabetically
listed according to their generic assignment, with
family level and higher classification omitted.
Proc. Linn. Soc. N.S.W., 131, 2010
Phylum Chordata Balfour, 1880
Class Conodonta Pander, 1856
Genus BELODINA Ethington, 1959
Type species
Belodus compressus Branson and Mehl, 1933.
Belodina compressa (Branson and Mehl, 1933)
Fig. 3A-B
Synonymy
Belodus compressus Branson and Mehl, 1933, p. 114,
pl. 9, figs 15, 16.
Belodus grandis Stauffer, 1935, p. 603-604, pl. 72,
figs 46, 47, 49, 53, 54, 57.
Belodus wykoffensis Stauffer, 1935, p. 604, pl. 72,
figs 51, 52, 55, 58, 59.
Oistodus fornicalus Stauffer, 1935, p. 610, pl. 75, figs
3-6.
Belodina dispansa (Glenister); Schopf, 1966, p. 43,
Ok I, ie, 7,
Belodina compressa (Branson and Mehl); Bergstr6m
and Sweet, 1966, p. 321-315, pl. 31, figs 12-19;
Sweet in Ziegler, 1981, p. 65-69, Belodina - plate
2, figs 1-4; Leslie, 1997, p. 921-926, figs 2.1-2.20,
3.1-3.4 (cum syn.); Zhen et al., 2004, p. 148, fig.
5A-I (cum syn.); Percival et al., 2006, fig. 3A-D.
Belodina confluens Sweet; Percival et al., 1999, p. 13,
Fig. 8.21.
Material
Ten specimens from two samples (see Table 1).
Discussion
Only compressiform (Fig. 3A) and grandiform
(Fig. 3B) elements were recovered from the
Tasmanian samples. These elements are identical with
those recorded from the upper part of the Wahringa
Limestone Member of the Fairbridge Volcanics
(assemblage C, see Zhen et al. 2004), and others from
drillcore samples in the Marsden district (Percival et
al. 2006) of central New South Wales. Morphological
distinction between B. compressa and closely related
species, particularly B. confluens, was discussed by
Zhen et al. (2004).
Genus CHIROGNATHUS Branson and Mehl, 1933
Type species
Chirognathus duodactylus Branson and Mehl,
1D)3;3}-
49
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 3. A-B, Belodina compressa (Branson and Mehl, 1933). A, compressiform element, AM F.136480,
JRC 2, inner-lateral view (1Y139-001); B, grandiform element, AM F.136481, JRC 2, outer-lateral view
(TY 139-003). C-F, Drepanodus sp. C, Sb element, AM F.136482, JRC 2, outer-lateral view (1Y 139-005).
D, Sb element, AM F.136483, JRC 2, inner-lateral view (1Y139-006). E, F, M element, AM F.136484, JRC
2, E, inner-lateral view ([Y139-004); F, basal view ([Y139-014). G-I, Gen. et sp. indet., all from YYF4,
G, Sc element, AM F.136485, inner-lateral view ([Y136-022); H, ?P element, AM F.136486, outer-lateral
view (IY 136-021); I, Sb element, AM F.136487, outer-lateral view (1Y136-019). J-K, Aphelognathus? sp.
from YYF4, J, Pb element, AM F.136488, inner-lateral view (LY 135-025). K, Pa element, AM F.136489,
inner-lateral view (1LY136-024). Scale bars 100 um.
revised the type species as having a seximembrate
or septimembrate apparatus, and concluded that
the 29 out of the 42 species recognized by Branson
and Mehl (1933), Stauffer (1935), and others since
the establishment of the genus could be confidently
Discussion
Chirognathus was established on 23 form species
recognized by Branson and Mehl (1933, pp. 28-34,
pl. 2) from the Harding Sandstone in Canyon City,
Colorado with Chirognathus duodactylus as the :
type species. Later Stauffer (1935) erected 15 form assigned to the genus, and in fact might belong to a
species of Chirognathus from the upper Glenwood _ Single species apparatus of his revised C. duodactylus.
Beds in the upper Mississippi Valley. Sweet (1982) He regarded 15 of Branson and Mehl’s (1933) and 13
50 Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
of Stauffer’s (1935) form species as junior synonyms
of C. duodactylus, with the M element represented by
form species C. duodactylus (= C. gradatus Branson
and Mehl, 1933, = C. planus Branson and Mehl,
1933), Sa by form species C. multidens Branson and
Mehl, 1933, Sb by form species C. panneus Branson
and Mehl, 1933 (= C. isodactylus Branson and Mehl,
1933), Sc by form species C. eucharis Stauffer, 1935,
Pa by form species C. varians Branson and Mehl,
1933 (= C. alternatus Branson and Mehl, 1933), and
Pb by form species C. monodactylus Branson and
Mehl, 1933 (= C. reversus Branson and Mehl, 1933).
As defined by Sweet (1982, p. 1039), C. duodactylus
has a ramiform-ramiform species apparatus including
a bipennate M element with a short and laterally
deflected anterior process and a long posterior process,
an alate Sa element with a straight, laterally extended
lateral process on each side, a digyrate Sb element
varying from subsymmetrical (with two processes
subequal in length) to markedly asymmetrical
. (with one lateral process longer than the other), a
bipennate Sc element with a shorter anterior process,
a bipennate Pa element resembling the Sc but with
the unit inwardly bowed with a more prominently
arched basal margin, and a digyrate Pb element with
two lateral processes directed in opposite directions
distally.
Chirognathus cliefdenensis Zhen and Webby,
1995, from the Cliefden Caves Limestone Subgroup of
central New South Wales, differs from C. duodactylus
in having distinctive blade-like P elements with high
processes bearing closely spaced, basally confluent
denticles (Zhen and Webby 1995, pl. 2, figs 13-16).
Chirognathus tricostatus sp. nov.
Figs 4-5
Synonymy
Chirognathus monodactyla Branson and Mehl;
Burrett, 1979, pp. 31-32.
Tasmanognathus careyi Burrett, 1979, p. 33-35,
partim, only pl. 1, fig. 12.
Derivation of name
Latin ¢ri- (three) and costatus (ribbed) referring to
the distinctive character, the tricostate cusp of the Sb,
Sc and Sd elements, of this Tasmanian species.
Material
71 specimens from eight samples (see Table 1).
Holotype: AMF.136496, YYF5, Sd element (Fig.
5A-C); paratypes: AM F.136490, C137c, Sa element
(Fig. 4A-C); AM F.136491, JRC 2, Sa element (Fig.
Proc. Linn. Soc. N.S.W., 131, 2010
AD); AM F.136492, YYF5, Sb element (Fig. 4E); AM
F.136493, C137c, Sb element (Fig. 4F); AM F.136494
(=UTG96872: Burrett 1979, pl. 1, fig. 12; originally
designated as one of the syntypes of T. careyi), Sb
element (Fig. 4G-H); AM F.136495, YYF5, Sc
element (Fig. 41-J); AM F.136497, C137c, Sd element
(Fig. SD-E); AM F.136498, C137c, Sd element (Fig.
5F-G); AM F.136499, JRC 2, Pa? element (Fig. 5H);
AM F.136500 (=UTG96866), JRC 2, Pa element
(Fig. 51); AM F.136501, JRC 2, Pa element (Fig. 5J-
K); AM F.136502, YYF1, Pb element (Fig. SL-N);
AM F.136503, YYF4, Pb element (Fig. 5O).
Diagnosis
A species of Chirognathus with a seximembrate
(possibly septimembrate) ramiform-ramiform
apparatus including alate Sa, modified digyrate Sb
and Sd, modified bipennate Sc, bipennate Pa and
digyrate Pb elements; all elements with long, peg-like
denticles, and a shallow, open basal cavity, typically
preserved without attachment of a basal funnel.
Description
Sa element symmetrical or nearly symmetrical,
with a prominent cusp and a denticulate lateral
process on each side (Fig. 4A-D); cusp large, straight,
antero-posteriorly compressed, with broadly convex
anterior and posterior faces and sharply costate
lateral margins; lateral processes extending laterally
and bearing three or more denticles of variable sizes,
which are also antero-posteriorly compressed; basal
cavity flared anteriorly and posteriorly with basal
margin nearly straight or slightly arched in posterior
or anterior view (Fig. 4A, D).
Sb element (Fig. 4E-H) like Sa, but asymmetrical
with outer lateral process slightly curved posteriorly
and with a short, but prominent costa developed
on the basal part of the anterior face (Fig. 4E, H);
outer lateral process slightly curved posteriorly and
also with basal margin twisted posteriorly and upper
margin anteriorly (Fig. 4G); basal cavity shallow,
flared anteriorly and posteriorly and extending
distally as a narrow and shallow groove underneath
each process (Fig. 4F).
Sc element modified bipennate, strongly asym-
metrical with denticulate anterior and posterior
processes and a strong costa on the outer lateral
face (Fig. 4I-J); both processes extending straight or
slightly curved inward; anterior process bearing three
or more denticles with the distal denticle (away from
the cusp) larger than the other denticles; posterior
process bearing two or more denticles with the
distal one (away from the cusp) larger than the other
denticle; larger denticle on the posterior or anterior
Sl
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 4. Chirognathus tricostatus sp. nov. A-D, Sa element; A-C, AM F.136490, paratype, C137c, A,
anterior view (LY 138-020), B, basal view ([Y138-021), C, posterior view (1Y142-023); D, AM F.136491,
paratype, JRC 2, anterior view (IY 142-002). E-H, Sb element; E, AM F.136492, paratype, YYF5, an-
terior view (IY135-039); F, AM F.136493, paratype, C137c, posterior view (LY138-022); G-H, AM
F.136494=UTG96872 (Burrett 1979, pl. 1, fig. 12; originally designated as one of the syntypes of T. careyi),
paratype, JRC 2, G, posterior view (IY141-018), H, anterior view (1Y141-019). I-J, Sc element, AM
F.136495, paratype, YYF5, I, upper-inner lateral view (1Y 135-035), J, upper-outer lateral view (LY 135-
036). Scale bars 100 um.
process being as wide as the cusp in the lateral view,
but more strongly compressed laterally than the cusp;
outer lateral costa prominent, forming a ridge-like
process near the base (Fig. 4J).
Sd element modified digyrate, strongly asym-
metrical with a robust cusp, a denticulate lateral
process on each side and a blade-like costa on the
anterior face (Fig. 5A-G); cusp tricostate with a sharp
costa along the lateral margins and on the broadly
convex anterior face, and a less convex posterior
face; anterior costa more strongly developed than that
in the Sb element, and extending to near the tip of the
cusp, and basally often developed into a short, blade-
a2
like process (Fig. 5C-D, G); lateral processes distally
curved posteriorly bearing three or more denticles of
variable sizes; basal cavity more open and strongly
flared posteriorly than that of the Sb element, forming
a strongly arched basal margin in posterior view (Fig.
SF).
Pa element bipennate with a prominent cusp and
denticulate anterior and posterior processes (Fig. 5H-
K); cusp suberect, laterally compressed with sharply
costate anterior and posterior margins and broadly
convex lateral faces (Fig. 5H-J); both anterior and
posterior processes bearing three or more denticles of
variable sizes, which are also laterally compressed;
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Figure 5. Chirognathus tricostatus sp. nov. A-G, Sd element; A-C, AM F.136496, Holotype, YYF5, A, an-
terior view (LY 135-034), B, posterior view (LY 142-028), C, upper view (LY135-033); D-E, AM F.136497,
paratype, C137c, D, anterior view (IY 142-025), E, posterior view (LY 138-027); F-G, AM F.136498, para-
type, C137c, F, posterior view (1Y 138-024), G, anterior view (LY 142-026). H, Pa? element; AM F.136499,
paratype, JRC 2, outer lateral view ([Y142-018); I-K, Pa element, I, AM F.136500 =UTG96866, para-
type, JRC 2, outer lateral view (TY141-026); J-K, AM F.136501, paratype, JRC 2, J, inner lateral view
(1Y142-020), K, basal view, close up showing the zone of recessive basal margin (LY 142-022). L-O, Pb
element; L-N, AM F.136502, paratype, YYF1, L, posterior view (LY 136-30), M, upper view (IY 136029),
N, anterior view (LY 142-029); O, AM F.136503, paratype, YYF4, basal-posterior view (LY 135-026). Scale
bars 100 um.
Proc. Linn. Soc. N.S.W., 131, 2010 53
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Fig. 6. A-I, Panderodus gracilis (Branson and Mehl, 1933). A, falciform, AM F.136504, JRC 2, outer-lat-
eral view (LY 139-033). B-C, truncatiform element, AM F.136505, JRC 2, B, posterior view (LY139-026);
C, inner-lateral view (LY 139-024). D-G, graciliform element; D-F, AM F.136506, JRC 2, D, inner-lateral
view (1Y139-017); E, outer-basal view of the basal part ([Y139-022); F, outer-lateral view (LY139-020);
G, AM F.136507, JRC 2, outer-lateral view ([Y139-023). H-I, falciform element; H, AM F.136508, JRC
2, inner-lateral view ([Y139-035); I, AM F.136509, YYF2, outer-lateral view ([Y140-25). J-N, Protopan-
derodus? nogamii (Lee, 1975). J, Sb element, AM F.136510, YYF4, outer-lateral view (1Y 136-027). K-N,
Pa element; K-L, AM F.136511, YYF4, K, outer-lateral view (1Y 136-025), L, outer lateral view, closer up
showing the furrow weaken and disappeared before researching basal margin ([Y136-026). M-N, AM
F.136512, YYF3, M, outer-lateral view ([Y 140-021), N, basal view (1Y140-019). Scale bars 100 1m unless
otherwise indicated.
anterior process typically slightly curved inward and
extending downward forming a gently arched basal
margin in lateral view (Fig. 5I-J); basal cavity shallow
and open, often with zone of recessive basal margin
preserved (Fig. 5K).
Pb element digyrate with a prominent cusp and
denticulate lateral process on each side (Fig. 5L-O);
cusp curved posteriorly with costate lateral margins;
lateral processes bearing four or more denticles of
variable sizes; basal cavity shallow and open with
54
gently arched basal margins in anterior or posterior
view (Fig. SL, N-O).
Discussion
Chirognathus tricostatus sp. noy. was initially
reported by Burrett (1979) as Chirognathus
monodactyla, one of the 23 form species recognized
by Branson and Mehl (1933). One of the syntypes of
Tasmanognathus careyi (AM F.136494 =UTG 96872)
is re-assigned herein to C. tricostatus to represent
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
the Sb position (Fig. 4G; also see Burrett 1979,
pl. 1, fig. 12). C. tricostatus from Tasmania differs
from two currently known multi-element species
of Chirognathus, C. duodactylus from the Upper
Ordovician (Sandbian) of North American Mid-
continent faunas and C. cliefdenensis from the Upper
Ordovician (Katian) of central New South Wales, in
having distinctive tricostate Sb, Sc and Sd elements.
Sweet (1982, 1988) recognized the M element
for the type species, C. duodactylus. A comparable
element has also recognized in the Tasmanian material
of C. tricostatus, but has been assigned to the Sd
position to form a symmetry transitional series with
other S elements. One of the illustrated specimens of
the Pa element (Fig. 5H) shows a nearly straight basal
margin and posteriorly curved cusp, and may possibly
represent the M element of this species. However, as
only one specimen is available in the current material,
it is tentatively assigned to the Pa element.
Genus PHRAGMODUS Branson and Mehl, 1933
Type species
Phragmodus primus Branson and Mehl, 1933.
Phragmodus undatus Branson and Mehl, 1933
Figs 7-8
Synonymy
Phragmodus undatus Branson and Mehl,
1933, p. 115-116, pl. 8, figs 22-26; Zhen
and Webby, 1995, p. 284, pl. 4, fig. 5; Leslie and
Bergstrém, 1995, p. 970-973, fig. 4.1-4.14 (cum
syn.); Zhen et al., 1999, p. 90, fig. 9.1-
9.5 (cum syn.); Zhen et al., 2003a, fig. 6N, O;
Pyle and Barnes, 2002, figs 14.11-14.12, 15.31-
15.32; Percival et al., 2006, fig. 4A-E.
Material
116 specimens from six samples (see Table 1).
Description
M element makellate, geniculate coniform with a
robust cusp and a short base triangular in outline (Fig.
7A-B); cusp strongly antero-posteriorly compressed
forming sharp lateral edges and broad anterior and
posterior faces; inner-lateral corner triangular in
outline, and outer-lateral proto-process short with
a gently arched upper margin; basal cavity shallow
with weakly wavy basal margins.
S elements ramiform bearing a long multi-
denticulate posterior process with one or two enlarged
denticles, but none of the Tasmanian specimens
Proc. Linn. Soc. N.S.W., 131, 2010
have the posterior process completely preserved. Sa
element symmetrical or nearly symmetrical with a
prominent costa on each side (Fig. 7C-D); posterior
process long with one denticle (typically the third
or fourth from the cusp) about twice as wide as the
adjacent denticles, and larger and longer than the
cusp; in some specimens a costa also developed
on each side of the larger denticle (Fig. 7D); basal
cavity shallow with strongly arched basal margins;
anterior (or antero-inner lateral) costa typically only
weakly developed (Fig. 7D). Sb element modified
quadriramate, like Sa but asymmetrical with the sharp
costate anterior margin curved inward (Fig. 7F-G).
Sc element modified bipennate, like Sb but strongly
asymmetrical with a sharply costate anterior margin
curved inward and with smooth inner and outer
lateral faces (Fig. 7H-L). Sd element tertiopedate,
like Sb, but with a broad anterior face and with one
of the larger denticles on the posterior process curved
inward and the other outward (Fig. 8A-C).
Pa element pastinate with long denticulate
posterior and inner lateral processes, and a suberect
cusp (Fig. 8D-G); cusp laterally compressed with
sharply costate anterior and posterior margins, outer
lateral face more convex; posterior process long,
bearing six or more denticles; inner lateral process
shorter, bearing five or more denticles and strongly
bending anteriorly forming an angle of nearly 180
degree with the posterior process (Fig. 8E, G); costate
anterior margin extending downward and not forming
a prominent anterior process (Fig. 8D); basal cavity
shallow, forming a wide and open groove along
the posterior and inner lateral processes, and flared
anteriorly and inner laterally (Fig. 8G). Pb element
pastinate, like Pa but with a more robust cusp and less
anteriorly curved inner lateral process (Fig. 8H-I).
Discussion
Leslie and Bergstr6m (1995) suggested a
seximembrate apparatus for P undatus, including
adenticulate makellate M, alate Sa, tertiopedate Sb,
bipennate Sc, pastinate Pa and Pb elements. All six
elements have been recovered from the Tasmanian
samples (Figs 7-8); they are identical with those
described and illustrated by Leslie and Bergstrém
(1995, fig. 4) from the Joachim Dolomite and Kings
Lake Limestone of Missouri, except that an additional
tertiopedate element was recognized in the Tasmanian
material (Fig. 8A-C). This latter element is similar
to the Sb element, but has the cusp and the larger
denticles on the posterior process strongly twisted
towards different sides in respect to the antero-
posterior axis. It is assigned herein to represent the
Sd position.
5
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Fig. 7. Phragmodus undatus Branson and Mehl, 1933. A-B, M element; A, AM F.136513, YYF4, posterior
view (1Y136-005); B, AM F.136514, YYF4, anterior view (1Y136-006). C-D, Sa element, AM F.136515,
C137c, C, basal view (LY 138-014); D, lateral view (LY138-015). E-G, Sb element; E, AM F.136516, YYF4,
outer-lateral view (LY 136-015); F, AM F.136517, YYF4, inner-lateral view (LY 136-014), G, AM F.136518,
YYF4, inner-lateral view (1Y136-016). H-L, Sc element; H, AM F.136519, YYF4, inner-lateral view
(LY 136-013); I-J, AM F.136520, YYF4, I, outer-lateral view (LY136-009), J, inner-lateral view (1Y136-
010); K-L, AM F.136521, C137c, K, basal view (LY 138-016), L, outer-lateral view (1Y 138-017). Scale bars
100 um.
56 Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Fig. 8. Phragmodus undatus Branson and Mehl, 1933. A-C, Sd element; AM F.136522, JRC 2, A, upper
view (IY 138-028), B, outer-lateral view (IY 138-029), C, posterior view (ITY 138-030). D-G, Pa element; D-
E, AM F.136523, YYF4, D, outer-lateral view ([Y136-001), E, basal view ([Y136-011); F-G, AM F.136524,
YYF4, F, inner-lateral view (LY136-003), G, basal view ([Y136-012). H-I, Pb element; H, AM F.136525,
YYF4, outer-lateral view (1Y¥136-004); I, AM F.136526, YYF4, antero-outer lateral view (1Y136-017).
Scale bars 100 um.
Genus PROTOPANDERODUS Lindstrém, 1971 Synonymy
Scolopodus nogamii Lee 1975, p. 179, pl. 2, fig. 13.
?Panderodus nogamii (Lee); Cantrill and Burrett
2004, p. 410, pl. 1, figs 1-16.
Panderodus nogamii (Lee); Zhang et al. 2004, p. 16,
pl. 5, figs 1-5.
Protopanderodus nogamii (Lee); Watson 1988: p.
124, pl. 3, figs 1, 6; Zhen et al. 2003b, p. 207-
Type species
Acontiodus rectus Lindstrém, 1955.
Protopanderodus? nogamii (Lee, 1975)
Fig. 6J-N
Proc. Linn. Soc. N.S.W., 131, 2010 57
LATE ORDOVICIAN CONODONTS FROM TASMANIA
209, fig. 23A-P, ?Q (cum syn.); Zhen and Percival
2004a, p. 104-105, fig. I8A-K (cum syn.).
Protopanderodus? nogamii (Lee); Zhen and Percival
2004b, p. 170-172, fig. 11P, Q (cum syn.).
Material
69 specimens from four samples (see Table 1).
Discussion
Recent review of this species by Cantrill and
Burrett (2004) suggested a geographical distribution
restricted to Gondwana and_peri-Gondwanan
terranes. Morphologically P nogamii is_ rather
conservative over its long stratigraphic range from
the upper Floian (evae Zone, Zhen et al. 2003b) to
upper Sandbian (uwndatus Zone, this study). Generic
assignment of this species has been debated in the
literature (see synonymy list). Most elements of this
species bear a non-panderodontid furrow on each side,
suggesting that it might be more closely related to
Protopanderodus rather than to typical Panderodus.
Genus TASMANOGNATHUS Burrett, 1979
Type species
Tasmanognathus careyi Burrett, 1979.
Diagnosis
Septimembrate apparatus with a ramiform-
pectiniform apparatus structure including makellate
M, ramiform S (including alate Sa with a denticulate
lateral process on each side, digyrate Sb, bipennate or
modified bipennate Sc, and tertiopedate Sd), carminate
Pa, and angulate Pb (some species with an additional
modified angulate or pastinate Pb2) elements.
Discussion
Following Burrett’s (1979, p. 32) original view
that Zasmanognathus might be closely related to
Rhipidognathus, Aldridge and Smith (1993) doubtfully
included it in the Rhipidognathidae. Affinities with
other genera remain conjectural, although greatest
similarities appear to be with Yaoxianognathus (see
discussion below).
Tasmanognathus was established on a single
species, 7. careyi Burrett, 1979 from the Lower
Member of the Benjamin Limestone in the Florentine
Valley and Everlasting Hills of central Tasmania.
Subsequently, Zasmanognathus has been reported
from the mid Darriwilian to upper Katian of eastern
Australia, North China (An et al. 1985, An and Zheng
1990, Pei and Cai 1987), Qinling Mountains in the
Kunlun-Qinling Region (Pei and Cai 1987), Tarim
58
Basin (Zhao et al. 2000; Jing et al. 2007), South
Korea (Lee 1982; An and Zheng 1990), ?Siberia
and northeastern Russia (Domoulin et al. 2002), and
possibly North America (where it was referred to
as Yaoxianognathus abruptus). It is represented by
nine named species and several additional unnamed
forms, the latter included herein in 7Zasmanognathus
although some are poorly known or inadequately
documented. Following is a brief review of the known
species (with our interpretation of element notations
in parentheses):
Tasmanognathus careyi Burrett, 1979 from the
Lower Limestone Member of the Benjamin Limestone
in the Florentine Valley and Everlasting Hills of
central Tasmania; a seximembrate apparatus was
originally recognized, but based on re-examination
of original topotypes and additional new material, it
has been revised herein as having an septimembrate
apparatus (including M, Sa, Sb, Sc, Sd, Pa, and Pb
elements).
Badoudus badouensis Zhang in An et al.,
1983 from the Fengfeng Formation (Sandbian) of
Handan, Hebei Province in North China (considered
by An et al. 1985, p. 102, to represent a species of
Tasmanognathus); this is a poorly defined form
species with only two specimens illustrated (An et al.
1983, pl. 25, figs 5, 6, text-fig. 12.17), both of which
are carminate, bearing an indistinctive cusp and a long
denticulate anterior process and a short denticulate
posterior process. This element is comparable with
the Pa element of Zasmanognathus defined herein.
Tasmanognathus borealis An in An et al., 1985
from the upper part of the Yaoxian Formation (late
Sandbian) of Yaozhou District (formerly Yaoxian) of
Tongchuan City, Shaanxi Province in North China;
originally defined as having a quinquimembrate
apparatus, including trichonodelliform (= Sa element;
see An et al. 1985, pl. 1, fig. 20), zygognathiform
(© Sbyelement: see An”etialsalOSse pkely fis. 13)
cordylodiform (= Sc element; see An et al. 1985, pl.
1, fig. 15), ozarkodiniform (= Pa element; see An et
al. 1985, pl. 1, fig. 14), and prioniodiniform (= Pb
element; see An et al. 1985, pl. 1, fig. 16).
Tasmanognathus gracilis An in An et al., 1985
from the upper part of the Yaoxian Formation (late
Sandbian) of Yaozhou District (formerly Yaoxian)
of Tongchuan City, Shaanxi Province in North
China; originally defined as having a seximembrate
apparatus, including cyrtoniodiform (= M element;
see An et al. 1985, pl. 1, fig. 8), trichonodelliform
(= Sa element; see An et al. 1985, pl. 1, fig. 12),
ligonodiniform (= Sb element; see An et al. 1985, pl.
1, fig. 11), cordylodiform (= Sc element; see An et al.
1985, pl. 1, fig. 10), ozarkodiniform (= Pa element:
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
see An et al. 1985, pl. 1, fig. 7), and prioniodiniform
(= Pb element; see An et al. 1985, pl. 1, fig. 9).
Tasmanognathus multidentatus An in An and
Zheng, 1990 (p. 20, 95, text-fig. 9, pl. 11, fig. 4); the
only figured specimen (pl. 11, fig. 4) is a Pa element
from the Yaoxian Formation of Yaozhou District
(formerly Yaoxian) of Tongchuan City, Shaanxi
Province in North China, which is identical with
the Pa element of 7. borealis An in An et al., 1985.
In fact, the figured Pa element (pl. 11, fig. 4) of 7
multidentatus and the holotype and a figured paratype
of T: borealis (An et al. 1985, pl. 1, figs 13, 16) were
recovered from the same sample (Tpl3y2). It is
unclear why An and Zheng (1990) tried to replace 7.
borealis with T: multidentatus. However, as the latter
is anomem nudum, T. borealis remains the valid name
for this Yaoxian species.
Tasmanognathus planatus Pei in Pei and Cai, 1987
from the Sigang Formation of Xichuan and Neixiang
Counties, Henan Province in the Qinling Mountains
. (Pei and Cai 1987; Chen et al. 1995; Wang et al.
1996); the type material was represented by Pa (Pei
and Cai 1987, pl. 13, fig. 12), Pb (Pei and Cai 1987,
pl. 13, figs 8, 713), and Sb (Pei and Cai 1987, pl. 13,
fig. 9) elements.
Tasmanognathus shichuanheensis An in An et al.,
1985 from the lower part of the Yaoxian Formation
(mid Sandbian) of Yaozhou District (formerly
Yaoxian) of Tongchuan City, Shaanxi Province
in North China; originally defined as having a
seximembrate apparatus, including cyrtoniodiform
GuvMmclementsisceAnwetale 1985s) pla tie. 3):
trichonodelliform (= Sa element; see An et al. 1985,
pl. 1, fig. 4), ligonodiniform (= Sb element; see An et
al. 1985, pl. 1, fig. 1), cordylodiform (= Sc element;
see An et al. 1985, pl. 1, fig. 5), ozarkodiniform (=
Pa element; see An et al. 1985, pl. 1, fig. 2), and
prioniodiniform (= Pb element; see An et al. 1985, pl.
1, fig. 6).
Tasmanognathus sigangensis Pei in Pei and Cai,
1987 from the Shiyanhe Formation (late Sandbian-
early Katian) of Neixiang County, Henan Province in
the Qinling Mountains; a quinquimembrate species
apparatus was recognized including trichnodelliform
(= Sa element; Pei and Cai 1987, pl. 13, fig. 4),
zygognathiform (= Sb element, Pei and Cai 1987,
pl. 13, fig. 11), cordylodontiform (= Sc element;
Pei and Cai 1987, pl. 13, fig. 7), prioniodiniform (=
Pa element; Pei and Cai 1987, pl. 13, figs 1-2), and
ozarkodontiform (= ?Pb element; Pei and Cai 1987,
joll, 13}, 10g, 3),
Tasmanognathus sishuiensis Zhang in An
et al. 1983 reported from the upper Fengfeng
Formation (early Sandbian) of Shandong and Hebei
Proc. Linn. Soc. N.S.W., 131, 2010
provinces in North China; defined as consisting
of a quinquimembrate apparatus including
trichonodelliform (= Sa element, see An et al.
1983, pl. 29, figs 7, 9, 10), zygognathiform (= Sb
element, see An et al. 1983, pl. 29, figs 4-6, 8, ?11),
cordylodontiform (= Sc element, see An et al. 1983,
pl. 29, figs 1-3), ozarkodiniform (= Pa element, see
Anetal. 1983, pl. 29, figs 14-15), and prioniodiniform
(= Pb element, see An et al. 1983, pl. 29, figs 12-13)
elements. This species is characterized by its widely
spaced peg-like denticles on the S elements.
Tasmanognathus sp. described by Pei and Cai
(1987) from the Sigang and Shiyanhe formations
of Neixiang County, Henan Province in the Qinling
Mountains; represented by cordylodontiform (=
Sc element; Pei and Cai 1987, pl. 13, figs 5-6) and
prioniodiniform (= ?Pb element; Pei and Cai 1987, pl.
13, fig. 10) elements.
Tasmanognathus sp. from the Fossil Hill
Limestone (early Katian) of the Cliefden Caves
Limestone Subgroup, central New South Wales was
only represented by the Pa element (Zhen and Webby
1995, p. 289, pl. 5, fig. 23), which showed close
resemblance to the Pa element of 7: borealis from the
Yaoxian Formation.
Tasmanognathus sp. cf. T. borealis An in An et
al., 1985; only the Pa element known from unnamed
limestone of Late Ordovician (late Sandbian) age
intersected in drillcore in the Marsden district of
south-central New South Wales (Percival et al.
2006).
The three species of Zasmanognathus (T. borealis,
T: gracilis and T. shichuanheensis) erected by Anin An
etal. (1985) from the Yaoxian Formation (Darriwilian-
Sandbian) of Yaozhou District (formerly Yaoxian) of
Tongchuan City, Shaanxi Province in North China
exhibit similar species apparatus and closely related
morphological variations of constituent elements. An
et al. (1985) established two conodont zones in the
Yaoxian Formation, namely the T. shichuanheensis
Zone in the lower part of the formation (Bed 1 to Bed
3, see An et al. 1985, fig. 2), and the Zasmanognathus
borealis-T. gracilis Zone spanning the upper part
of the Yaoxian Formation (Bed 4 to Bed 8) into the
basal part of the overlying Taoqupo Formation (Bed
9). An and Zheng (1990, p. 95, text-fig. 9) suggested
that 7: sishuiensis from the Fengfeng Formation
might be the direct ancestor of the species from the
Yaoxian Formation, and indicated an inferred lineage
from 7. sishuiensis to T: shichuanheensis and then
to T. multidentatus (= T. borealis). They showed the
morphological changes of the three species, mainly
from widely spaced denticles on the processes of the
S and Pb elements and a prominent cusp on the Pa
SY
LATE ORDOVICIAN CONODONTS FROM TASMANIA
element of 7: sishuiensis, to closely spaced denticles in
the S and Pb elements and an indistinctive cusp inthe Pa
element of 7. mu/tidentatus (= T. borealis). However,
these species from the Yaoxian Formation and
Fengfeng Formation show some detailed differences
in composition of the apparatus in comparison with
T. careyi from Tasmania. In particular, they seem to
lack makellate M and tertiopedate Sd elements, and
have a “dolabrate” Sc element with a long denticulate
posterior process. Morphologically, such features
support a closer relationship with Yaoxianognathus
yaoxianensis An in An et al., 1985. However, these
species lack hindeodellid denticles on the processes
of the S elements, which was the major character
that An (in An et al. 1985) employed to distinguish
Yaoxianognathus from Tasmanognathus. As revision
of An’s species of Zasmanognathus from the Yaoxian
Formation and the Fengfeng Formation of North
China is beyond the scope of the current study, they
are retained in Zasmanognathus for the time being,
although they show some significant differences in
morphology and apparatus composition in comparison
with the type species of Zasmanognathus as revised
here.
Based on the concept of Yaoxianognathus
employed by An (in An et al. 1985) and others (e.g.
Savage 1990; Zhen et al. 1999), generic assignment of
species previously included in Yaoxianognathus but
which apparently lack hindeodellid denticles on the
processes of the S elements, should be reconsidered.
For example, Yaoxianognathus abruptus (Branson
and Mehl, 1933), a North American Midcontinent
species ranging across the wndatus to tenuis zones
of the Mohawkian, was initially proposed as a form
species based only ona carminate Pa element (Branson
and Mehl, 1933, pl. 6, fig. 11) and revised by Leslie
(2000, p. 1143) as having a seximembrate apparatus.
It closely resembles An’s species of Tasmanognathus
from North China; most importantly, none of Leslie’s
illustrated S elements of ¥. abruptus (fig. 4.15-4.18)
bears hindeodellid denticles that are characteristic of
Yaoxianognathus, and hence we suggest this species
more likely belongs to Tasmanognathus.
Similarly, S elements of Yaoxianognathus?
neonychodonta Zhang, Barnes and Cooper, 2004,
from the Stokes Siltstone of the Amadeus Basin in
central Australia, lack hindeodellid denticles and
therefore should be excluded from Yaoxianognathus.
As Zhang et al. (2004) implied, this species may be
more closely related to Plectodina, judging from
the morphological characters of its ramiform S and
pastinate Pb elements.
In comparison, the two multielement species
of Yaoxianognathus from the Upper Ordovician of
60
central New South Wales (¥. wrighti Savage, 1990 and
Y. ani Zhen, Webby and Barnes, 1999) do exhibit well
developed hindeodellid denticles on the processes
of the S elements, particularly on the long posterior
process of the Sc element (Savage 1990, fig. 6.7-6.12;
Zhen et al. 1999, fig. 15.3-15.6, 15.9-15.12, 15.16).
The apparatuses of both species include a makellate
M and a modified bipennate Sc elements, which differ
morphologically from corresponding elements in the
T. careyi apparatus as defined herein.
Tasmanognathus careyi Burrett, 1979
Figures 9-15
Synonymy
Tasmanognathus careyi Burrett, 1979, p. 33-35,
partim only text-figs 2-4, pl. 1, figs 1-7, 11, 13-19
(text-fig. 2 = Pb2 , text-fig. 3 = Pa, text-fig.4A =
Sb, text-fig. 4B = Sc, text-fig. 4C, D = Sd; pl. 1,
figs 1-3 = Pb2, 4-5 = Pbl, 6-7 = Pa, fig. 11 = Sc,
figs 13-14= Sb, figs 15-18 = Sd, fig. 19 = Sa); non
fig. 12 = C. tricostatus sp. nov., non figs 8-10, 20
=, sp. chcareyi.
? Tasmanognathus careyi Burrett; An and Zheng,
LO9Os pl alalchiesy 2)
Material
297 specimens from nine samples (see Table 1).
Burrett (1979, p. 33, pl. 1, figs 1-7, 11-12, 17-18,
20) designated 11 figured specimens from sample
JRC 2 as syntypes, ten of which (excluding UTG
96863 which was not able to be located for this
study; figured by Burrett 1979, pl. 1, figs 17-18),
and 225 additional specimens (including originally
undesignated topotypes) from five samples (LLMB,
C137, C98, JRC 2 and JRB, see Table 1) are available
for the current study. AM F.136547 (~UTG 96851;
Burrett 1979, pl. 1, fig. 6) representing a Pa element
is selected herein as lectotype (Fig. 14A-B); and
seven out of ten originally designated and illustrated
syntypes were examined and illustrated herein as
paralectotypes, including AM F.136557 (=UTG
96857, Fig. 15H; Burrett 1979, pl. 1, fig. 1), AM
F.136559 (-UTG 96860; Fig. 15K; Burrett 1979,
pl. 1, fig. 2), AM F.136560 (=UTG 96853, Fig. 15L;
Burrett 1979, pl. 1, fig. 3), AM F.136553 (~UTG
96850, Fig. 15A-B; Burrett 1979, pl. 1, fig. 4), AM
F.136554 (=UTG 96882, Fig. 15C; Burrett 1979, pl.
1, fig. 5), AM F.136548 (~UTG 96856, Fig. 14C;
Burrett 1979, pl. 1, fig. 7), and AM F.136539 (~UTG
96876, Fig. 12A-C; Burrett 1979, pl. 1, fig. 11).
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Figure 9. Tasmanognathus careyi Burrett, 1979. M element; A, AM F.136527 =UTG96875, JRC 2, ante-
rior view (1Y141-025). B, AM F.136528, C137c, anterior view ([Y138-011). C-D, AM F.136529, C137c,
C, posterior view (1Y 138-008); D, basal view (TY138-009). E-F, AM F.136530, JRC 2, E, posterior view
(LY 138-035); F, basal view (1Y138-034). G-J, AM F.136531, JRC 2, G, upper view (1Y139-007); H, pos-
terior view (LY139-008); I, inner-lateral view (LY139-009); J, anterior view (1Y139-010). Scale bars 100
um.
Proc. Linn. Soc. N.S.W., 131, 2010 61
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 10. Tasmanognathus careyi Burrett, 1979. Sa element; A-B, AM F.136532 =UTG96874 (Burrett
1979, pl. 1, fig. 19), JRC 2, A, posterior view (TY 141-022), B, lateral view (LY141-021); C-E, AM F.136533,
YYF5, C, anterior view (1Y140-004), D, posterior view ([Y140-003), E, upper-posterior view (1Y140-
001); F-I, AM F.136534, YYF4, F, lateral view (TY135-019), G, posterior view (1Y135-020), H, anterior
view (LY135-018), I, upper view, close up showing the cross section of the cusp ([Y 135-022). Scale bars
100 um.
UTG 96877, previously designated as a syntype
(Burrett 1979, pl. 1, fig. 20) is excluded from this
species and re-assigned to T. sp. cf. careyi representing
the Sb position (AM F.136567, Fig. 16G herein).
Another previously designated syntype UTG 96872
(Burrett 1979, pl. 1, fig. 12) is also excluded from this
species and re-assigned to Chirognathus tricostatus
sp. nov. where it represents the Sb position (AM
F.136494, Fig. 4G-H herein).
Diagnosis
Septimembrate apparatus with a ramiform-
pectiniform structure including makellate M, alate Sa,
digyrate Sb, bipennate Sc, tertiopedate Sd, carminate
Pa, angulate Pbl, and pastinate Pb2 elements. S
elements with a robust cusp, an open and shallow
basal cavity, and long closely-spaced denticles on the
processes; Pa element with a longer anterior process,
62
a nearly straight basal margin and a cusp varying
from prominently larger (juvenile) than adjacent
denticles to rather indistinctive in size (when mature).
Pb1 element with a robust cusp, and a strongly curved
basal margin. Pb2 element with a short adenticulate
outer lateral process, long denticulate anterior and
posterior processes, and a strongly laterally flared
base.
Description
M element makellate with a denticulate inner-
lateral process bearing three to five pointed denticles
(Fig. 9), and a shorter, typically adenticulate outer
lateral process (Fig. 9A-C, H); cusp robust, antero-
posteriorly compressed (Fig. 9G), with a sharp costa
along the inner-lateral and outer lateral margins (Fig.
9G-I), and distally curved posteriorly (Fig. 9C-D,
G); denticles on the inner lateral process also antero-
Proc. Linn. Soc. N.S.W., 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Figure 11. Jasmanognathus careyi Burrett, 1979. Sb element; A, AM F.136535 =UTG96873, posterior
view (1Y141-027); B-D, AM F.136536, C98, B, basal view ([Y137-039), C, anterior view (1Y137-037),
D, basal-posterior view (LY137-038); E, AM F.136537, C98, outer-anterior view (LY 137-034); F-H, AM
F.136538 =UTG96898 (Burrett, 1979, fig. 4A), C98, F, anterior view (TY137-031), G, upper view (IY 137-
032), H, upper-posterior view (LY 137-030). Scale bars 100 um.
posteriorly compressed, with a sharp costa along the
inner-lateral and outer-lateral margins (Fig. 9C, G-
H); basal cavity shallow and open, tapering towards
distal ends of the processes and flaring posteriorly
(Fig. 9D, F), and often with weakly developed zone of
recessive basal margins (Fig. 9F); anterior portion of
basal margin nearly straight (Fig. 9B, J), but posterior
portion weakly curved (Fig. 9C, E, H).
Sa element alate (Fig. 10), symmetrical with a
robust cusp, a prominent tongue-like anticusp, and
a long denticulate lateral process on each side; cusp
proclined, subquadrate in cross section (Fig. 10E,
I), with a sharp costa on each side (Fig. 10A-B) and
often a weak costa along the posterior margin (Fig.
10D-E), but some specimens with a broad posterior
face (Fig. 10G) or with a broad carina developed
(Fig. 10A); broad anterior face bearing a shallow but
prominent mid groove and a broad carina on each
side (Fig. 10C, H); cusp extended downward to form
a downward extending tongue-like anticusp (Fig.
10A-D, H); lateral process long, bearing up to ten or
more closely spaced denticles (Fig. 10C-D, H), which
Proc. Linn. Soc. N.S.W., 131, 2010
are compressed antero-posteriorly; basal cavity open
and shallow, flared posteriorly; basal margin arched
in posterior view (Fig. 10D).
Sb element digyrate, asymmetrical, with a robust
cusp, long denticulate process on each side, and
a prominent downwardly extending tongue-like
anticusp (Fig. 11); cusp suberect, slightly curved
inward (Fig. 10A), with a more strongly convex
anterior face, and a sharp costa on each side (Fig.
11G-H); outer lateral process shorter, bearing three or
more denticles (Fig. 11A, D, G); inner lateral process
longer, bearing five or more peg-like denticles (Fig.
11C, F), and more strongly curved posteriorly (Fig.
11B, G), forming an angle of about 100-110 degrees
between the two processes in the upper or basal view
(Fig. 11B, G).
Sc element bipennate, asymmetrical with a robust
cusp, a long denticulate posterior process, and a
short denticulate anterior process (Fig. 12); cusp
suberect basally and reclined distally (Fig. 12A, F, H)
with a more convex outer lateral face, and laterally
compressed with a sharp costa forming anterior
63
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 12. Tasmanognathus careyi Burrett, 1979. Sc element; A-C, AM F.136539 =UTG96876 (Burrett
1979, pl. 1, fig. 11), JRC 2, paralectotype, A, inner lateral view (1Y141-020), B, basal view (LY 141-016), C,
outer lateral view (LY141-015); D-E, AM F.136540, YYF5, D, inner-basal view (LY 135-041); E, inner-lat-
eral view (LY 135-040); F-G, AM F.136541, JRC 2, F, inner lateral view (LY 139-028); G, inner-basal view
(TY 139-027); H, AM F.136542, YYFS5, inner-lateral view (1Y140-009); I-J, AM F.136543 =UTG 96899
(Burrett, 1979, fig. 4B), C98, I, inner-lateral view ([Y137-029), J, basal view (LY137-027). Scale bars 100
pm.
and posterior margins (Fig. 12F-I); anterior margin
curved inward (Fig. 12D-I); posterior process bearing
three or more (up to seven) denticles, which are
laterally compressed and posteriorly reclined (Fig.
11A, C, I); anterior process with upper margin curved
inwards, and extending downwards bearing two to
four small denticles (Fig. 12D-H); basal cavity open
and shallow, slightly flared inwards (Fig. 12B, D, G),
some specimens with basal funnel attached (Fig. 12I-
I:
Sd element tertiopedate, weakly asymmetrical to
nearly symmetrical with a robust cusp, a prominent
anticusp, a denticulate posterior process and a
denticulate lateral process on each side (Fig. 13);
cusp with a broad anterior face (Fig. 13B-C), and
with a prominent costa along the posterior margin and
on each lateral side (Fig. 13D, G); anticup short and
downward extending (Fig. 13C-D); posterior process
64
long and straight, broken in most specimens, in one
of the examined specimens bearing ten denticles (Fig.
13A); lateral process bearing four or more denticles
(Fig. 13B-D); basal cavity open, T-shaped in basal
view (Fig. 13B).
Pa element carminate (Fig. 14), laterally
compressed and blade-like, with a small cusp, and
with the anterior and posterior processes bearing
basally confluent denticles; cusp erect (smaller
specimens, Fig. 14C, E) to slightly inclined (larger
specimens, Fig. 14A, D), typically larger and higher
than adjacent denticles (Fig. 14C, E, F), but less
distinctive in the larger specimens (Fig. 14 A, H);
two processes of unequal length, anterior process
longer and higher, bearing five to eight closely-
spaced denticles; posterior process lower and shorter,
bearing two to six denticles, with distal end slightly .
bent downward (Fig. 14A, D); juvenile specimens
Proc. inn. Soe. NSW. 131, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Figure 13. Tasmanognathus careyi Burrett, 1979. Sd element; A, AM F.136544 =UTG96902 (Burrett,
1979, pl. 1, figs 15-16), LLMB, upper view (I1Y137-033); B-E, AM F.136545 =UTG96900 (Burrett, 1979,
fig. 4C-D), B, basal view (1Y137-023), C, anterior view (LY 137-022), D, lateral view ([Y127-024), E, close
up showing fine striae on the surface of the cusp (1Y137-026); F-G, AM F.136546, YYF1, G, lateral view
(LY140-015), F, basal-posterior view (LY 140-014). Scale bars 100 um.
Figure 14. Tasmanognathus careyi Burrett, 1979. Pa element; A-B, AM F.136547 -UTG96851 (Burrett
1979, pl. 1, fig. 6), lectotype, JRC 2, A, outer lateral view ([Y141-002), B, basal-inner lateral view (1Y141-
003); C, AM F.136548 =UTG96856 (Burrett 1979, pl. 1, fig. 7), paralectotype, JRC 2, outer lateral view
(1Y 141-004); D, AM F.136549 =UTG96893a (Burrett, 1979, fig. 3), LLM (B), inner-lateral view (LY 137-
001); E, AM F.136550 =UTG96893b (Burrett, 1979, fig. 3), LLM (B), outer-lateral view (LY 137-003); F-
G, AM F.136551, C98, F, outer-lateral view (1Y137-005), G, upper view (LY 137-006); H-I, AM F.136552,
JRC 2, H, inner-lateral view (LY 137-010), I, basal view (LY137-009). Scale bars 100 pm.
Proc. Linn. Soc. N.S.W., 131, 2010 65
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 15. Tasmanognathus careyi Burrett, 1979. A-G, Pb1 element; A-B, AM F.136553 =UTG96850
(Burrett 1979, pl. 1, fig. 4), paralectytype, JRC 2, A, basal-outer lateral view (1Y141-007), B, outer lat-
eral view (1Y141-006); C, AM F.136554 =UTG96882 (Burrett 1979, pl. 1, fig. 5), paralectotype, JRC 2,
inner lateral view (1Y 141-008); D-E, AM F.136555, C98, D, basal view (LY137-018), E, inner lateral view
(TY 137-019); F-G, AM F.136556, C98, F, outer-lateral view (LY 137-021), G, basal view (1Y 137-020). H-M,
Pb2 element; H, AM F.136557 =UTG96857 (Burrett 1979, pl. 1, fig. 1), paralectotype, JRC 2, outer lat-
eral view (TY 141-009); I-J, AM F.136558, JRC 2, I, inner-lateral view (1Y137-040), J, basal view (TY 137-
041); K, AM F.136559 =UTG96860 (Burrett 1979, pl. 1, fig. 2), paralectotype, JRC 2, outer lateral view
(TY 141-011); L, AM F.136560 =UTG96853 (Burrett 1979, pl. 1, fig. 3), paralectotype, JRC 2, outer lateral
view (1Y141-012); M, AM F.136561, JRC 2, outer-lateral view (1Y137-014). Scale bars 100 um.
66 Proc. Linn. Soc. N.S.W., 131, 2010
VNC ZEEN Cs BURREM: KGSPERCIWAL AND Bays LIN
exhibiting a prominently lower and shorter posterior
process with two to four less closely-spaced denticles
(Fig. 14C, E); basal cavity shallow and open, flared
laterally and extended toward distal end of the
processes as a tapering shallow groove (Fig. 141);
basal margin nearly straight to slightly arched in
lateral view (Fig. 14A, C-E, F).
Pbl1 element angulate (Fig. 15A-G), laterally
compressed and blade-like, with a robust cusp, and
denticulate anterior and posterior processes; cusp
strongly compressed laterally, more convex outer
laterally, suberect and slightly curved inwards
with sharp anterior and posterior margins; two
processes typically sub-equal in length (Fig. 15E)
or with slightly longer posterior process (Fig. 15B,
C), bearing three to six short, laterally compressed
and basally confluent denticles; anterior process
extending downward forming an angle of about 100-
120 degrees between the two processes in lateral view
(Fig. 15B, F); basal cavity shallow and open, laterally
. flared and extended as a shallow groove underneath
each process (Fig. 15D, G).
Pb2 element pastinate (likely a variant of the Pb1
element), with a robust cusp, long denticulate anterior
and posterior processes, and a short adenticulate outer
lateral process (Fig. 15H-M); cusp suberect (Fig.
15H), laterally compressed, with a broad smooth inner
lateral face, and sharp anterior and posterior margins,
outer lateral face smooth (Fig. 15L-M) or with a
mid costa (Fig. 15H, K); anterior process typically
longer, bearing up to seven or more denticles, which
are typically closely spaced with confluent bases
(Fig. 151, M); most specimens with posterior process
broken, bearing up to five denticles (Fig. 15L); outer
lateral process typically represented by a prominent
tongue-like basal extension (Fig. 15J-M), or as a short
adenticulate process (Fig. 15H); basal cavity shallow,
outer laterally flared more strongly, and tapering as
a shallow groove to the distal end of anterior and
posterior processes (Fig. 15I-J).
Discussion
One originally designated syntype (AM F.136567
=UTG96877 Fig. 16G; also see Burrett 1979, pl.
1, fig. 20) and an additional figured specimen (AM
F.136562 =UTG96904, Fig. 16A; also see Burrett
1979, pl. 1, figs 8-10) of T. careyi are excluded from
this species and re-assigned to represent the Sb and
Pb2 elements of T. sp. cf. careyi, as they exhibit more
widely spaced denticles on the processes.
The original definition of the S element given by
Burrett (1979) is more or less followed herein, except
that his Sal element (Burrett 1979, fig. 4C-D) is now
assigned to the Sd position (Fig. 13), the digyrate
Proc. Linn. Soc. N.S:W., 131, 2010
element with a longer inner lateral process (Burrett
1979, fig. 4A, referred to as Sc) to the Sb position
(Fig. 11), and the bipennate element with a shorter
downwardly extended and inner laterally curved
anterior process (Burrett 1979, fig. 4B, referred to
as Sb) to the Sc position (Fig. 12). Burrett (1978, p.
34) further recognized an Sa2 element with the cusp
exhibiting a subquadrate cross section, but illustrated
it as Sa (Burrett 1979, pl. 1, fig. 19; also Fig. 10A-
B herein). This symmetrical or nearly symmetrical
element (Fig. 10) is confirmed as occupying the Sa
position. The makellate M element (Fig. 9) described
herein was not recognized in Burrett’s original
description of TZ careyi. Specimens originally
included in the Pa element by Burrett (1979) show
two morphotypes, which are defined herein to
represent the Pb! (Burrett 1979, pl. 1, figs 4-5; Fig.
1SA-G) and Pb2 (Burrett 1979, pl. 1, figs 1-3; Fig.
15H-M) elements. They can be easily differentiated
from each other by having a short tongue-like outer
lateral process, a costa on the outer lateral face or
a short adenticulate outer lateral process in the Pb2
element (Fig. 15H-M).
Burrett (1979, pp. 33-34) discussed the
considerable ontogenetic variations among the P
elements, in particular the posterior process of the
Pa element (referred to as the Pb element by Burrett,
1979, see p. 33, fig. 3) and the anterior process of
the Pb2 element (assigned to part of the Pa element
by Burrett, 1979, see p. 33, fig. 2). Juveniles of the
Pa element have a larger cusp and a lower posterior
process with less closely spaced denticles (Fig.
14C, E; Burrett 1979, fig. 3). It cannot presently be
established whether the distinctions between the
Pb1 and Pb2 elements, and within the Pa elements,
represent ecophenotypic variations, or whether they
reflect a high degree of morphological plasticity.
T. careyi has been widely reported from North
China (Zhao et al. 1984; Wang and Luo 1984; Pei and
Cai 1987; An and Zheng 1990). However, judging
from the illustrations of these specimens, none can be
confidently assigned to the Tasmanian species, except
for one specimen figured by An and Zheng (pl. 11, fig.
2) from the lower part of the Yaoxian Formation in the
Ordos Basin of Shaanxi Province that is comparable
with the Pa element of 7: careyi. Pa elements of 7:
borealis (An in An et al. 1985, pl. 1, fig. 14) and 7.
multidentatus (An and Zheng 1990, pl. 11, fig. 4; =
T. borealis), also from the Yaoxian Formation of the
Ordos Basin, similarly have an indistinct cusp that
is nearly the same size as adjacent denticles, but the
outline of these two illustrated specimens is shorter
and higher in comparison with the Pa element of 7.
careyi (Fig. 14).
67
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Figure 16. Tasmanognathus sp. cf. careyi Burrett, 1979. A, Pb2 element, AM F.136562 =UTG96904 (Bur-
rett, 1979, pl. 1, figs 8-10), LLM (B), outer-lateral view (1Y 137-007). B, M element, AM F.136563, YYF1,
posterior view ([Y136-032). C-F, Sa element; C-D, AM F.136564, JRC 2, C, Posterior view (LY138-006);
D, postero-basal view (LY 138-005); E, AM F.136565, JRC 2, posterior view (LY 139-029); F, AM F.136566,
YYF1, anterior view (1Y140-010). G, Sb element, AM F.136567 =UTG96877 (syntype of T. careyi; Bur-
rett 1979, pl. 1, fig. 20), JRC 2, posterior view (1Y141-024). H-I, Sd element, AM F.136568, YYF4, H,
posterior view (IY 135-023), I, upper view (LY135-024); J-K, Sc element, AM F.136569, YYF1, J, inner
lateral view (1Y140-013), K, outer lateral view (1Y140-012). Scale bars 100 um.
Tasmanognathus sp. cf. T. careyi Burrett, 1979
Figures 16-17
Synonymy
Tasmanognathus careyi Burrett, 1979, p. 33-35,
partim only pl. 1, figs 8-10 (= Pb2 element), fig.
20 (= Sb element).
Material
34 specimens from four samples in the Settlement
Road section of Florentine Valley area (see Table 1).
68
Diagnosis
A species of Zasmanognathus
septimembrate apparatus, including makellate M, alate
Sa, digyrate Sb, bipennate Sc, digyrate? (modified
tertiopedate) Sd, carminate Pa, angulate? (bipennate)
Pb1, and pastinate Pb2 elements; elements robust and
large in size bearing a prominent cusp ornamented
with fine striae, and small widely spaced denticles on
the processes of M, S, Pb1 and Pb2 elements; most
elements with basal funnel attached.
having an
Proc. Linn. Soe. N.SOW.a3 1, 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Fig. 17. Tasmanognathus sp. cf. T. careyi Burrett, 1979. A-H, Pa element; A-C, AM F.136570, YYF4,
A, inner-lateral view (LY135-005), B, upper view (LY135-003), C, outer-lateral view, close up showing
fine surface striae (1Y135-007); D-F, AM F.136571, YYF4, D, outer-lateral view (1Y135-009), E, basal
view (1Y135-008), F, outer-lateral view, close up showing rounded boring hole on the surface ([Y135-
010); G-H, AM F.136572, YYF4, G, inner lateral view (1Y 140-037), H, close up showing fine surface
striae ([Y140-038). I-K, Pb1 element; I, AM F.136573, YYF4, outer-lateral view (1Y 135-012); J-K, AM
F.136574, YYF4, J, inner-lateral view (LY 140-031), K, outer lateral view (LY 140-030). Scale bars 100 pm
unless otherwise indicated
Description
M element with a long, denticulate inner-lateral
process bearing five short and widely spaced denticles
(Fig. 16B), and a short, outer lateral process bearing
two small rudimentary denticles; cusp robust, antero-
posteriorly compressed with a sharp costa along the
iner-lateral and outer lateral margins and distally
curved posteriorly.
Sa element alate (Fig. 16C-F), with a robust cusp
and a long denticulate lateral process on each side;
cusp strongly compressed antero-posteriorly, with a
sharp costa along the lateral margins; lateral process
long, bearing three or more peg-like denticles (Fig.
16C), which are also strongly compressed antero-
posteriorly, basal cavity open and shallow, flared
posteriorly, isosceles-triangular in basal view (Fig.
Proc. Linn. Soc. N.S.W., 131, 2010
16D-E); basal margin gently arched in posterior view
(Fig. 16C).
Sb element digyrate, like Sa but asymmetrical
(Fig. 16G); cusp robust and antero-posteriorly
compressed with sharp lateral margins; denticulate
lateral process on each side bearing two or three short
widely-spaced denticles; inner lateral process longer
and more downwardly extending.
Sc element bipennate, strongly asymmetrical
with a robust cusp, denticulate anterior and posterior
processes (Fig. 16J-K); cusp distally curved inner
laterally with a more convex outer lateral face bearing
a prominent costa; posterior process longer and
slightly arched bearing three widely-spaced denticles;
anterior process curved inward bearing two widely-
spaced denticles.
69
LATE ORDOVICIAN CONODONTS FROM TASMANIA
Sd element digyrate? with a robust cusp, a long
denticulate lateral process on each side, a sharp costa
on the posterior face and a broad anterior face with a
weak carina (Fig. 16H-I); cusp with a sharp costa on
each side and on the posterior face, and ornamented
with fine striae; inner lateral process longer bearing
eight small denticles.
Pa element blade-like with a prominent cusp,
and denticulate anterior and posterior processes
(Fig. 17A-H); cusp suberect or slightly inclined
posteriorly, laterally compressed, standing higher
above the adjacent denticles, and about twice width
of the adjacent denticles on the anterior process, and
typically leaving a prominent notch between cusp and
the first denticle on the anterior process (Fig. 17A,
G); anterior process higher and longer bearing four to
eight larger and basally confluent denticles (Fig. 17A,
D, G); posterior process slightly shorter, triangular
in outline in lateral view, with a tapering distal end
and bearing five or six smaller denticles (Fig. 17A,
G); basal cavity shallow, flared laterally, forming a
shallow groove underneath each process (Fig. 17E),
and with a straight basal margin (Fig. 17D); some
specimens bearing fine rounded boring holes (Fig.
17F).
Pb1 element asymmetrical with a suberect, robust
cusp and long denticulate anterior and posterior
processes (Fig. 17I-K); cusp curved inward, diamond-
shaped in cross section with a sharp costa along the
anterior and posterior margins, a mid costa on the
inner lateral face (Fig. 17J), and a broad carina on the
outer lateral face (Fig. 17K); two processes bearing
small, discrete denticles; posterior process longer
with six or more denticles, and anterior process
shorter, extending downwards (Fig. 17K).
Pb2 element pastinate, with a robust cusp and
denticulate anterior, posterior and outer lateral
processes (Fig. 16A); cusp laterally compressed with
a sharp costa along anterior and posterior margins and
on the outer lateral face; long anterior and posterior
processes bearing short, widely spaced denticles;
outer lateral process short, represented by a single
denticle.
Discussion
This species differs from 7. careyi in having a
Pa element with shorter and higher outline bearing
a prominent cusp and a notch in front of the cusp,
and in having the S, Pb1 and Pb2 elements bearing
small, discrete or widely-spaced denticles on the
processes. Additional specimens from the Settlement
Road section of Florentine Valley area confirm that
it represents a separate species of Zasmanognathus.
However, as only a small number of specimens are
70
available for study, this species is retained herein
under open nomenclature pending further collecting
and study.
ACKNOWLEDGMENTS
Field work in Tasmania by Zhen was supported by a
grant from the Betty Mayne Scientific Research Fund of the
Linnean Society of New South Wales. Burrett’s study was
funded by the Australian Research Council. Gary Dargan
(Geological Survey of New South Wales) assisted with
acid leaching and residue separation. Scanning electron
microscope photographs were prepared in the Electron
Microscope Unit of the Australian Museum. We thank
Stephen Leslie and John Pickett for their perceptive and
constructive reviews of the manuscript. The study was
undertaken by Zhen as part of a CAS/SAFEA International
Partnership Program for Creative Research Teams,
and is a contribution to IGCP Project 503: Ordovician
Palaeogeography and Palaeoclimate. Percival publishes
with permission of the Director of the Geological Survey of
New South Wales.
REFERENCES
Aldridge, R.J. and Smith, M.P. (1993). Conodonta. 563-
572. In Benton, M.J. (ed.), ‘The Fossil Record 2’. 845
p. (Chapman and Hall, London).
An, T.X., Zhang, F., Xiang, W.D., Zhang, Y.Q., Xu, W.H.,
Zhang, H.J., Jiang, D.B., Yang, C.S., Lin, L.D., Cui,
Z.T. and Yang, X.C. (1983). ‘The conodonts in North
China and adjacent regions’. 223 p. (Science Press:
Beijing) (in Chinese with English abstract).
An, T.X., Zhang, A.T. and Xu, J.M. (1985). Ordovician
conodonts from Yaoxian and Fuping, Shaanxi
Province, and their stratigraphic significance. Acta
Geologica Sinica 59, 97-108 (in Chinese with English
abstract).
An, T.X. and Zheng, S.C. (1990). ‘The conodonts of the
marginal areas around the Ordos Basin, North China’.
199 pp. (Science Press: Beijing) (in Chinese with
English abstract).
Banks, M.R. and Burrett, C.F. (1980). A preliminary
Ordovician biostratigraphy of Tasmania. Journal of the
Geological Society of Australia 26, 363-375.
Balfour, F. M. (1880-1881). A treatise on comparative
embryology. Two volumes. Macmillan & Co., London.
Bergstrom, S.M. and Sweet, W.C. (1966). Conodonts
from the Lexington Limestone (Middle Ordovician)
of Kentucky and its lateral equivalents in Ohio and
Indiana. Bulletin of American Paleontology 50 (229),
271-441.
Branson, E.B. and Mehl, M.G. (1933). Conodont studies.
University of Missouri Studies 8, 1-349.
Proc: mn: Soe sNsSeW St 2010
Y.Y. ZHEN, C.F. BURRETT, I.G. PERCIVAL AND B.Y. LIN
Burrett, C.F. (1978). Middle-Upper Ordovician conodonts
and stratigraphy of the Gordon Limestone Sub-group,
Tasmania. Unpublished PhD thesis, University of
Tasmania, 342p.
Burrett, C.F. (1979). Tasmanognathus: a new Ordovician
conodontophorid genus from Tasmania. Geologica et
Palaeontologica 13, 31-38.
Burrett, C.F., Banks, M.R., Clota, G. and Seymour, D.
(1989). Lithostratigraphy of the Ordovician Gordon
Group, Mole Creek, Tasmania. Records of the Queen
Victoria Museum, Launceston 96, 1-14.
Burrett, C.F., Long, J. and Stait, B. (1990). Early-Middle
Palaeozoic biogeography of Asian terranes derived
from Gondwana. In “Palaeozoic Palaeogeography and
Biogeography’ (eds McKerrow, W.S. and Scotese, C.
R.). Geological Society Memoir 12, 163-174.
Burrett, C.F., Stait, B.A., and Laurie, J. (1983). Trilobites
and microfossils from the Middle Ordovician of
Surprise Bay, southern Tasmania, Australia. Memoir
of the Australian Association of Palaeontologists 1,
177-193.
Burrett, C.F., Stait, B-A., Sharples, C. and Laurie, J.
(1984). Middle-Upper Ordovician shallow platform to
deep basin transect, southern Tasmania, Australia. In
“Aspects of the Ordovician System’ (ed D.L. Bruton)
Universitetsforlaget, Oslo, 149-158.
Cantrill, R.C. and Burrett, C.F. (2004). The greater
Gondwana distribution of the Ordovician conodont
Panderodus nogamii (Lee) 1975. Courier
Forschungsinstitut Senckenberg 245, 407-419.
Chen, X., Rong, J.Y., Wang, X.F., Wang, Z.H., Zhang,
Y.D. and Zhan, R.B. (1995). Correlation of the
Ordovician rocks of China: charts and explanatory
notes. International Union of Geological Sciences,
Publication 31, 1-104.
Corbett, K.D. and Banks, M.R. (1974). Ordovician
stratigraphy of the Florentine Synclinorium south
west Tasmania. Papers and Proceedings of the Royal
Society of Tasmania 107, 207-238.
Domoulin, A. G., Harris, A. G., Gagiev, M., Bradley,
D. C. and Repetski, J.E. (2002). Lithostratigraphic,
conodont, and other faunal links between lower
Paleozoic strata in northern and central Alaska and
northeastern Russia. Geological Society of America,
Special Paper 360, 291-312.
Ethington, R.L. (1959). Conodonts of the Ordovician
Galena Formation. Journal of Paleontology 33, 257-
292.
Goodwin, P.W. and Anderson, E.J. (1985). Punctuated
Aggradational Cycles: a general hypothesis of episodic
stratigraphic accumulation. Journal of Geology 93,
515-533.
Jing, X.C., Yang, Z.L., Zhang, F., Zhang, S.B. and Deng,
S.H. (2007). Conodont biostratigraphy of Ordovician
outcrop section in Kalpin region of the Tarim Basin,
Xinjiang, China. Acta Palaeontologica Sinica 46
(Suppl.), 201-207.
Laurie, J.R. (1991). Articulate brachiopods from the
Ordovician and Lower Silurian of Tasmania. Memoir
of the Australian Association of Palaeontologists 11,
1-106.
Proc. Linn. Soc. N.S.W., 131, 2010
Lee, H.Y. (1975). Conodonten aus dem unteren
und mittleren Ordovizium von Nordkorea.
Palaeontographica Abteilung A 150, 161-186.
Lee H. Y. (1982). Conodonts from the Hoedongri
Formation (Silurian), western Jeongseon area,
Kangweon-do, South Korea. Journal of the National
Academy of Sciences, Republic of Korea, Natural
Sciences 21, 43-131.
Leslie, S.A. (1997). Apparatus architecture of Belodina
(Conodonta): Interpretations based on fused clusters of
Belodina compressa (Branson and Mehl, 1933) from
the Middle Ordovician (Turinian) Plattin Limestone
of Missouri and lowa. Journal of Paleontology 71,
921-926.
Leslie, S.A. (2000). Mohawkian (Upper Ordovician)
conodonts of eastern North America and Baltoscandia.
Journal of Paleontology 74 (6), 1122-1147.
Leshe, S.A. and Bergstrém, S.M. (1995). Element
morphology and taxonomic relationships of the
Ordovician conodonts Phragmodus primus Branson
and Mehl, 1933, the type species of Phragmodus
Branson and Mehl, 1933, and Phragmodus undatus
Branson and Mehl, 1033. Journal of Paleontology 69
(5), 967-974.
Lin Baoyu and Qiu Hongrong (1990). Geological and
geographical distribution of the Ordovician conodont
genus Zasmanognathus Burrett in North China.
Courier Forschungsinstitut Senckenberg (CFS) 117,
55-59.
Lindstr6ém, M. (1955). Conodonts from the lowermost
Ordovician strata of south-central Sweden. Geologiska
Foreningens i Stockholm Foerhandlinger 76, 517-604.
Lindstrém, M. (1971). Lower Ordovician conodonts of
Europe. In “Symposium on conodont biostratigraphy’
(eds W.C. Sweet and S.M. Bergstré6m). Geological
Society of America, Memoir 127, 21-61.
Pander, C. H. (1856). “Monographie der fossilen
Fische des Silurischen Systems der Russisch-
Baltischen Gouvernements’. 91 p. (Akademie der
Wissenschaften, St. Petersburg).
Pei, F. and Cai, S.H. (1987). “Ordovician conodonts from
Henan Province’. 128 p. (Press of the Wuhan College
of Geosciences, Wuhan). (in Chinese)
Percival, I.G., Morgan, E.J. and Scott, M.M. (1999).
Ordovician stratigraphy of the northern Molong
Volcanic Belt: new facts and figures. Geological
Survey of New South Wales, Quarterly Notes 108, 8-
Dip
Percival, I.G., Zhen, Y.Y. and Pickett, J.W. (2006). Late
Ordovician faunas from the Quandialla-Marsden
district, south-central New South Wales. Proceedings
of the Linnean Society of New South Wales 127, 235-
DSS;
Pyle, L.J. and Barnes, C.R. (2002). “Taxonomy, evolution,
and biostratigraphy of conodonts from the Kechika
Formation, Skoki Formation, and Road River Group
(Upper Cambrian to Lower Silurian), Northeastern
British Columbia’. 227 p. (NRC Research Press,
Ottawa).
Savage, N.M. (1990). Conodonts of Caradocian (Late
Ordovician) age from the Cliefden Caves Limestone,
71
LATE ORDOVICIAN CONODONTS FROM TASMANIA
southeastern Australia. Journal of Paleontology 64,
821-831.
Schopf, T.J. (1966). Conodonts of the Trenton Group
(Ordovician) in New York, southern Ontario, and
Quebec. New York State Museum and Science Service
Bulletin 405, 1-105.
Stait, B. (1984). Ordovician nautiloids of Tasmania -
Gouldoceratidae fam. nov. (Discosorida). Proceedings
of the Royal Society of Victoria, 96 (4), 187-207.
Stait, B. (1988). Tasmanian nautiloid faunas
— biostratigraphy, biogeography and morphology.
Senckenbergia Lethaia 69, 87-107.
Stait, B. and Flower, R.H. (1985). Michelinoceratida
(Nautiloidea) from the Ordovician of Tasmania,
Australia. Journal of Paleontology 59 (1), 149-159.
Stauffer, C.R. (1935). Conodonts of the Glenwood beds.
Bulletin of Geological Society of America 46, 125-168.
Sweet, W.C. (1982). Conodonts from the Winnipeg
Formation (Middle Ordovician) of the northern Black
Hills, South Dakota. Journal of Paleontology 56,
1029-1049.
Sweet, W.C. (1988). ‘The Conodonta: Morphology,
Taxonomy, Paleoecology, and Evolutionary History
of a Long-Extinct Animal Phylum’. 212 p. (Clarendon
Press, Oxford).
Teichert, C. and Glenister, B.F. (1952). Fossil nautiloid
faunas from Australia. Journal of Paleontology 26,
730-752.
Wang X.F., Chen, X., Chen, X.H. and Zhu, C-Y. (1996).
“Stratigraphical Lexicon of China, the Ordovician
System’. 126p. (Geological Publishing House,
Beijing) (in Chinese).
Wang, Z.H. and Luo, K.Q. (1984). Late Cambrian and
Ordovician conodonts from the marginal areas of the
Ordos Platform, China. Bulletin, Nanjing Institute of
Geology and Palaeontology, Academia Sinica 8, 237-
304.
Watson, S.T. (1988). Ordovician conodonts from the
Canning Basin (W. Australia). Palaeontographica
Abteilung A 203 (4-6), 91-147.
Webby, B.D., VandenBerg, A.H.M., Cooper, R.A., Banks,
M.R., Burrett, C.F., Henderson, R.A., Clarkson, P.D.,
Hughes, C.P., Laurie, J., Stait, B., Thomson, M.R.A.
and Webers, G.F. (1981). The Ordovician System in
Australia, New Zealand and Antarctica. Correlation
Chart and Explanatory Notes. /nternational Union of
Geological Sciences, Publication No. 6, 1-64.
Zhang, J.H., Barnes, C.R. and Cooper, B.J. (2004).
Early Late Ordovician conodonts from the Stokes
Siltstone, Amadeus Basin, central Australia. Courier
Forschungsinstitut Senckenberg 245, 1-37.
Zhao, S.Y., An, T.X., Qiu, H.R., Wan, S.L. and Ding, H.
(1984). ‘Palaeontological atlas of North China III,
Micropalaeontology’. 857 p. (Geological Publishing
House, Beijing) (in Chinese).
Zhao, Z.X., Zhang, G.Z. and Xiao, J.N. (2000). “Paleozoic
stratigraphy and conodonts in Xinjiang’. 340 p.
(Petroleum Industry Press, Beijing) (in Chinese with
English Abstract).
72
Zhen, Y.Y. and Percival, 1.G. (2004a). Middle Ordovician
(Darriwilian) conodonts from allochthonous
limestones in the Oakdale Formation of central New
South Wales, Australia. Alcheringa, 28, 77-111.
Zhen, Y.Y. and Percival, I1.G. (2004b). Middle
Ordovician (Darriwilian) conodonts from the
Weemalla Formation, south of Orange, New South
Wales. Memoirs of the Association of Australasian
Palaeontologists, 30, 153-178.
Zhen, Y.Y., Percival, I.G. and Farrell, J.R. (2003a). Late
Ordovician allochthonous limestones in Late Silurian
Barnby Hills Shale, central western New South Wales.
Proceedings of the Linnean Society of New South
Wales 124, 29-51.
Zhen, Y.Y., Percival, I.G. and Webby, B.D. (2003b). Early
Ordovician conodonts from far western New South
Wales Australia. Records of the Australian Museum
55, 169-220.
Zhen, Y. Y., Percival, I1.G. and Webby, B.D. (2004).
Conodont faunas from the Mid to Late Ordovician
boundary interval of the Wahringa Limestone Member
(Fairbridge Volcanics), central New South Wales.
Proceedings of the Linnean Society of New South
Wales 125, 141-164.
Zhen, Y.Y. and Webby, B.D. (1995). Upper Ordovician
conodonts from the Cliefden Caves Limestone
Group, central New South Wales, Australia. Courier
Forschungsinstitut Senckenberg 182, 265-305.
Zhen, Y.Y., Webby, B.D. and Barnes, C.R. (1999).
Upper Ordovician conodonts from the Bowan Park
succession, central New South Wales, Australia.
Geobios 32, 73-104.
Ziegler, W. (ed.) (1981). “Catalogue of Conodonts, Vol.
4’. 445 p. (Schweizerbart’sche Verlagsbuchhandlung,
Stuttgart).
Ziegler, W. (ed.) (1991). ‘Catalogue of Conodonts, Vol.
5°. 212 p. (Schweizerbart’sche Verlagsbuchhandlung,
Stuttgart).
Proc. Linn. Soc. N.S.W., 131, 2010
Stratigraphic Revision of the Hatchery Creek Sequence (Early-
Middle Devonian) Near Wee Jasper, New South Wales
JAMES R. HUNT AND GAVIN C. YOUNG
Research School of Earth Sciences, The Australian National University, Canberra ACT 0200, Australia
Ghunt595@gmail.com) (Gavin. Young@anu.edu.au)
Hunt, J.R., and Young, G.C. (2010). Stratigraphic revision of the Hatchery Creek sequence (Early-Middle
Devonian) near Wee Jasper, New South Wales. Proceedings of the Linnean Society of New South Wales
131, 73-92.
A new formation (the Corradigbee Formation) is erected for the upper part of the previous ‘Hatchery Creek
Conglomerate’, which is elevated to Group status, its lower part renamed the Wee Jasper Formation. The
“Hatchery Creek Conglomerate’, south of Burrinjuck Dam and 50 km northwest of Canberra, was previously
defined as a 2.9 km thick sedimentary sequence of conglomerate, sandstone and shale nonconformable on
underlying Lower Devonian limestones. The coarser lower part (Wee Jasper Formation) is now estimated at
about 1500 m thick; an additional type section is nominated for its upper part, which was not included in the
original type section, and lithologies, subdivision, and contacts with underlying and overlying formations
are described. The upper sequence of dark shales and mudstones (Corradigbee Formation) has an estimated
thickness of about 260 m, with 15 fining-upward cycles in which 50 new fossil sites have been found.
Repetition of lower strata of the Hatchery Creek sequence in the west, due to an unrecognised syncline
axis through the central part of the outcrop area, had suggested a much greater thickness than interpreted
in this study. The relatively high topography of the softer shales and mudstones in the core of the syncline
is a transient inverted topography resulting from recently eroded Tertiary basalts. The whole sequence is
interpreted as conformable on underlying limestones, and of Emsian-Eifelian age.
Manuscript received 30 October 2009, accepted for publication 17 February 2010.
KEYWORDS: Corradigbee Formation, Emsian-Eifelian, Hatchery Creek Group, Wee Jasper Formation.
INTRODUCTION
The previously named ‘Hatchery Creek
Conglomerate’ is a thick sedimentary sequence of
Devonian non-marine strata located 5Okm NW of
Canberra (Fig. la). It is exposed over an area of about
70 km?, with most of its outcrop on the Brindabella
1:100 000 sheet, about 4 km? of which is covered by
remnant Tertiary basalt (Owen and Wyborn 1979),
and a small northern extension on the Yass 1:100
000 sheet (Cramsie et al. 1978). Underlying marine
limestones of the Murrumbidgee Group, in the
Goodradigbee valley near the village of Wee Jasper
(Fig. 1b), contain an abundant invertebrate fauna,
including conodonts, brachiopods, and corals (see
Pedder et al. 1970, and references therein). These
provide a late Early Devonian (Emsian) maximum
age limit for the Hatchery Creek sequence.
The ‘Hatchery Creek Conglomerate’ was
originally assumed to be Upper Devonian in age,
based on lithological similarity with the Hervey Group
of central New South Wales (Pedder 1967, Conolly,
in Packham 1969, Pedder et al. 1970). However a
fossil fish assemblage discovered during geological
mapping by Owen and Wyborn (1979) was described
by Young and Gorter (1981) as probably late Eifelian
(Middle Devonian) in age.
Previous authors, when referring to the
‘Hatchery Creek Conglomerate’, commented
on the most accessible lower section, formed
predominantly of cycles of massive conglomerate
and sandstone. The measured section of Owen and
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
149
v Yass
| setae. @,. Burrinjuck Dam 3
a | JX f)
yy Study se A \
\
[\
Gundagai
¢
q Tumut
y=
Lake George
50km
Type Section
[upper Wee Jasper Fm]
__Type Section
[lower Wee Jasper Fm]
-(Owen & Wyborn 1981)
Inferred syncline axis
\(Hood & Durney 2002)
:
. Wee Jasper
[_|corradigbee Fmn ell Hatchery
)
[= |Wee Jasper Fmn (WJF Creek Gp
[| Murrumbidgee Gp Limestones
Burrinjuck Granite Complex
DEVONIAN
Early - Middle
Peppercorn ——a
“ |Beds 1km
Middle |
Figure 1. a. Regional locality map showing the study area. b. Generalised geological map showing the
outcrop area of the Hatchery Creek Group, based on the Owen and Wyborn (1979) Brindabella 1:100
000 geological map, updated by detailed field mapping (e.g. eastern areas of basalt; large area to west not
remapped). Previous fossil localities are the original fish locality at Windy Top (WT) described by Young
and Gorter (1981), and a second fish-plant locality (JF) studied by Francis (2003). The syncline axis as
identified in this study (on the left) is compared with the position of this structure inferred by Hood and
Durney (2002). Boxed study areas are shown in more detail in Figs. 2-4 as indicated.
74 Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
Wyborn (1979) did not reach into the upper sequence
above the lower massive conglomerates (Figs. 1b, 2a).
The fossil fish assemblage of Young and Gorter (1981)
occurs within the upper finer sequence of siltstones
and mudstones, in which almost no conglomeratic
horizons are seen. In this paper this upper sequence
is separated out as the new Corradigbee Formation,
described below, and the lower coarser sequence is
renamed the Wee Jasper Formation, both formations
included in the Hatchery Creek Group.
A second fossil locality (plants) was recorded on
the geological map of Owen and Wyborn (1979). In
1988 an ANU student excursion located fish remains
about 4 km south of the original fossil fish locality
(locality 59, Fig. 3a), and apparently higher in the
sequence. However the faunal composition seemed
identical to that from the original fish locality,
suggesting problems with the stratigraphy and
structure. The plant locality of Owen and Wyborn
(1979) was investigated by Francis (2003), where
_ fish were found in association, this locality (JF,
Figs. 1b, 2a, 3a, 4b, 5a) being only slightly higher
in the sequence than the original fish locality, now
called ‘Windy Top’ (WT, Fig. 1b). Hunt (2005,
2008) conducted a detailed field study of the upper
fine-grained sequence (Corradigbee Formation), and
discovered many additional fossil localities (Fig.
3a), mainly fish and plant remains, but with a few
invertebrates (gastropods, and probable arthropods;
see Appendix). New fish taxa in these assemblages
(Table 1) include several osteichthyans (bony fish),
and a new placoderm genus probably belonging to
the arthrodires (Hunt and Young, in press; Young et
al. 2010, fig. 4A). Fifteen fining-upward sedimentary
cycles were identified, comprising about 260 m of the
Corradigbee Formation. The cycles were mapped on
both sides of the axis of a broad syncline, a major
structure not shown on the geological map of Owen
and Wyborn (1979). As a result their estimated total
thickness of at least 2900 m for the entire sequence is
erroneous. The results presented here conform closely
with the first geological investigation of the area, in
an unpublished honours thesis by Edgell (1949).
The original fish locality was estimated at about
1.9 km above the base of the sequence, and it was
suggested that any disconformity with the underlying
limestones was of short duration (Owen and Wyborn
1979; Young and Gorter 1981). Previously, Edgell
(1949) had interpreted a conformable boundary
between the Hatchery Creek sequence and the
underlying limestones, an interpretation now followed
here (see below).
Physiographically, the Hatchery Creek area of
outcrop is part of the ‘Bimberi-Brindabella Upland’
Proc. Linn. Soc. N.S.W., 131, 2010
of Owen and Wyborn (1979, fig. 5), across which
Miocene basalts spread into the mapped area from
the ‘Kiandra Tableland’. The higher relief of the
softer mudstone sequence in the “middle ridge’ of the
mapped area of Hunt (2005, 2008; Fig. 3a) probably
results from inverted topography. It coincides with the
syncline axis, the topographic expression of which
has evidently been masked by recent erosion of the
cover of Tertiary basalt. Probably the basalt flowed
down a previous valley representing the eroded core
of the syncline, the basalt cover then inhibiting further
erosion until it was eventually stripped off. A small
residual cap of basalt remains adjacent to the original
fossil fish locality at ‘Windy Top’ (~700 m elevation,
Fig. 1b), with larger outcrops 3-5 km to the south and
west (Owen and Wyborn 1979). A flagstone quarry
at about 760 m elevation is located in the basalt that
forms the highest part of the middle ridge of the
mapped area, including Goodradigbee Hill (803 m;
Fig. 3a). The area of finer sedimentary rocks was
cleared for grazing many years ago, in contrast to the
timbered ridges to the east in the coarser sandstone
and conglomerates lower in the Hatchery Creek
sequence, but since completion of this study has been
revegetated as plantation pine forest.
Original access to the main outcrop was up the
Cave Creek Road (locked from 2008) and along the
‘Main Ridge Trail’ to the north, then west along the
“Windy Top Trail’ to the original fish locality. Access
to ‘Corradigbee’ homestead (Fig. 3a) is off the access
road to the 330kv power transmission line, from the
south via the Tumut Road.
METHODS
Reconnaissance mapping of the lower part of
the Hatchery Creek sequence by Young (1969) has
been reinvestigated during many excursions to collect
fossils following the research of Young and Gorter
(1981), and associated with the honours project of
Francis (2003). The detailed study of Hunt (2005)
involved about 30 days field work on the Corradigbee
Formation, covering about 20 km* in the upper section
of the Hatchery Creek sequence (rectangle, Fig. 1b).
The softer mudstone sequence is deeply eroded by
two north-flowing tributaries of MacPhersons Swamp
Creek, here termed ‘eastern creek’ and ‘western
creek’, separated by the prominent ‘middle ridge’
(Fig. 3a). Erosion gullies give many good exposures
of the softer sediments, and improved exposure
and accessibility was a result of the 2003 bushfires
in the Wee Jasper area, which burnt blackberry
infestations.
15
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
a b
LEGEND a
> ~<& endpoints of type wi z
sections Bi
Xsee=ssX measured section os
a ped etm tracks 25 25 2 4j
wee eee = faults ou J
°
WT, JF [J fossil fish sites
A Windy Top
w
>
WEE JASPER FORMATION to Pee
Murr. G
limesto1
8
Windy Top Trail
Type Section
po UY ISL
hho
Cave Creek Rd
Type Section
Proc. mn. soc. Nissw. 1B, 2010
J.R. HUNT AND G.C. YOUNG
Only some of the more significant fossil material
collected from many new localities has been prepared
and identified. The original description (Young
and Gorter 1981) documented such forms as the
placoderm Sherbonaspis hillsi (Fig. 2c), which closely
resembled the ‘winged fish’ first described by Hugh
Miller (1841) from classic Middle Devonian Old Red
Sandstone fish faunas of Scotland. This was the first
discovery of such an assemblage from the Southern
Hemisphere. An updated faunal list for the Hatchery
Creek fish assemblage is given in Table 1; formal
fossil descriptions will be presented elsewhere.
For the Corradigbee Formation, various field
sites were examined as to the bedding type, dip, strike,
lithology and sedimentary structures (see Appendix).
Many fining-upward sedimentary cycles could be
seen on air photographs by their more resistant
basal sandstones, and were traced out on a 90x90
cm photo enlargement. Some identified beds were
walked along strike to establish correlations between
_ different exposures for the detailed stratigraphy (Figs.
2a, 4a). Sedimentary strata with good exposure were
selected for measured stratigraphic sections using
either a tape or 150 cm Jacobs staff and abney level.
The cycle containing the original 1981 fossil fish
locality (WT) was called Cycle A, with overlying
cycles labelled up through the sequence as B, C, etc.,
and underlying cycles down the sequence labelled B’-
F’. The thickness of the Wee Jasper Formation was
estimated using aerial photographs and data plotted
from the lowest beds of the Corradibee Formation
and measured off the maps and photos.
Numbered localities are shown in Fig. 3aand listed
in the Appendix. For different field investigations the
locality numbers are: 1-24, 59-159 (Hunt 2005); 160-
161, 062-082 (Hunt 2008); prefix GY (Young 1969);
prefix JF (Francis 2003). All grid references refer to
the Wee Jasper 1:25 000 topographic map 8627-4N
(second edition, 2003). Full grid references (as in
appendix) are abbreviated in the text (e.g. 646385
611805 shortened to GR46385 1805). Fossil material
is registered in the ANU palaeontological collection,
Canberra (Building 47, Research School of Earth
Sciences).
PREVIOUS STRATIGRAPHY
The ‘Hatchery Creek Conglomerate’, named by
Joplin etal. (1953), consists of cyclothems of terrestrial
conglomerates, sandstones and mudstones. These fine
upwards and the beds are laterally extensive, some
being traceable over several kilometres along the
length of the outcrop (Young 1969). These beds can
be classified as red beds according to the definition of
Van Houten (1973).
Owen and Wyborn’s (1979) estimated thickness
of about 2.9 km for the Hatchery Creek Conglomerate
was followed by other authors (Young and Gorter 1981;
Branagan and Packham 2000; Packham 2003). With
the subdivision of this sequence into two formations
as proposed here (the Wee Jasper Formation and the
Corradigbee Formation), and the recognition that
the previously interpreted upper ~300 m of coarse
sandstones and conglomerates is in fact a repetition
of the lower strata (Wee Jasper Formation) on the
western limb of a syncline, a significantly reduced
total thickness estimate of 1760 m for the Hatchery
Creek Group is based on the following: thickness for
the lower formation (Wee Jasper Formation) estimated
from air photos (average dip 40°) at about 1500 m;
thickness for the upper Corradigbee Formation (as
defined below) estimated at 260 m.
HATCHERY CREEK GROUP (UPGRADED
FROM FORMATION)
WEE JASPER FORMATION (NEW NAME)
The first published description (as “Hatchery
Creek Conglomerate’) recorded numerous fining-
upward conglomeratic cycles (Owen and Wyborn
1979: microfiche M314-M320). A type section
comprising about 1200 m of almost continuous
exposure of cycles of ‘conglomerate, sandstone and
siltstone typical of the lower part of the formation’
was nominated along the Cave Creek Road (see
Fig. 1b), from the basal contact with the underlying
carbonates at their stated grid reference (GRS509
176), to the top at the T-junction of the Cave Creek
Figure 2 (LEFT). a. Detailed geological map of the Wee Jasper Formation (previously Hatchery Creek
Conglomerate, lower part) between the original type section (Cave Creek Road) for the lower part de-
fined by Owen and Wyborn (1979), and the new type section for the upper part (Windy Top Trail) de-
scribed in the text. Coarser basal part of each fining-upward unit indicated by stippling or shading. b.
Summary section for the lower 1600 m of the Hatchery Creek Group, showing correspondence between
the upper cycles of the Wee Jasper Formation and lower cycles of the Corradigbee Formation. c. Recon-
struction of the placoderm fish Sherbonaspis hillsi Young and Gorter (1981), which established a prob-
able Eifelian age for the Hatchery Creek sequence.
Proc. Linn. Soc. N.S.W., 131, 2010
VW
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
Road and Main Ridge Trail (their GR491 172; Fig.
2a). Owen and Wyborn (1979) noted a change at
about 1500 m above the base of the formation to a
lithology dominated by fine buff sandstone and red
siltstone with root casts. They considered but did not
follow the stratigraphic subdivision first proposed
by Edgell (1949), who separated off this finer upper
sequence as the ‘Middle Ridge Shales’ from the lower
‘Wee Jasper Creek Conglomerates’ (also overlooked
by Packham 1969; Pedder et al. 1970).
Young (1969) had previously subdivided the
lower 1550 m of the Hatchery Creek Conglomerate
into four units, the lower Units | and 2 forming the
eastern slope of the main ridge along the western
margin of the Goodradigbee valley, and the upper Units
3 and 4 mainly outcropping in the western drainage of
Macphersons Swamp Creek. The top of the formation
was left undifferentiated. This subdivision has been
checked in the field since 2003, supported by air
photo interpretation using new colour air photos, and
more recently Google Earth images, as summarised
in Figure 2a. Estimated thickness from the base for
these four units was 250, 200, 400 and 700 m (Young
1969). Owen and Wyborn (1979) stated that the cycles
as defined by the beds of conglomerate rarely extend
beyond about | km, but some of the units mapped
by Young (1969), for example the prominent basal
conglomerates of Units | and 2, can be traced on air
photos nearly 10 km along the western escarpment
of the Goodradigbee valley (Fig. 2a). The basal
conglomerates of Unit 2 form a row of conspicuous
outcrops about one third of the distance up the slope
of each spur between about GR495 210 and GR492
220. Both horizons can be traced north (with two
slight fault displacements at about GR495 222 and
GR492 232) at least to GR490 245. Unit 3 crops out
near the top and over the ridge to the west.
To the south, prominent outcrops of three ridges
north of the road in the Cave Creek Road type section
of Owen and Wyborn (1979) can be assigned to the
basal coarse beds of Units 1-3 (between GR509 174
and 504 171). The basal conglomerate of Unit 3 can
be readily traced on air photos from GY52 (GR499
193) to a prominent knoll on the spur at GR497
197, and then to the crest of the main ridge between
GR492 208 and 489 219. Farther north a sharp bend
to the west in the track crosses the basal conglomerate
of Unit 4 at GR4855 221. This basal conglomerate
is readily traced along strike to the south as a series
of prominent outcrops between valleys (e.g. GR487
2125, 487 208), and forms the first outcrop of
conglomerate encountered after the turnoff into the
eastern end of the Windy Top track, at GR489 2015.
Since the existing type section finishes well below
78
the lithological change to much finer sediments (the
base of our new formation), we nominate an additional
type section for the upper part of the renamed Wee
Jasper Formation, along the Windy Top Trail from
its junction with the main track at GR491 201, to
the vicinity of the locked gate at Windy Top (GR477
2016), about 1.4 km to the west. This is accessible
by 4-wheel drive vehicle, and the valleys to the north
and south display a thick section of alternating coarse
and fine beds as mapped by Young (1969). From the
eastern end of this type section, down the spurs into
the Goodradigbee valley, air photos clearly show the
base of Unit 3 at GR494 201, the base of Unit 2 at
GR496 2065, and the base of the Hatchery Creek
Group (and Unit | of the Wee Jasper Formation) on
the edge of the treeline at GR5012 202.
Owen and Wyborn (1979) recorded a fine-
grained sequence between about 1500-2600 m above
the base of their Hatchery Creek Conglomerate,
then a return to cyclic conglomerates about 300 m
thick at the top of the sequence. However our more
detailed mapping has shown this interpretation to be
incorrect, these ‘upper’ conglomerate cycles in fact
representing a repetition of the contact between the
Wee Jasper Formation and the Corradigbee Formation
on the western limb of the syncline. The western
contact (running beneath the largest basalt outcrop;
Fig. 1b) was not mapped in detail, but approximates
to the corresponding formation boundary of Edgell
(1949). The most westerly discovered fossil site
(Fig. 3a, locality 160; with fish and plants) is still
in the Corradigbee Formation. Further west, light
yellow sandstones of the Wee Jasper Formation were
observed in the vicinity of GR449 174, but to the
north similar horizons are more conglomeratic where
they emerge from beneath the basalt (near GR450
203). A similar increase in coarseness to the north
was observed on the eastern limb of the syncline
(see below). The uppermost coarse layers of the Wee
Jasper Formation are exposed within the main outcrop
of the Corradigbee Formation, in the creek bed along
a section of the Western Creek (dashed line, Fig. 5a),
but too narrow to be shown on the geological map
(Fig. 1b). Here, the lower levels of the Corradigbee
Formation beneath measured section 2 (see Fig. 3)
are inaccessible with a steep drop down to the creek
bed.
Lower and Upper Contacts
Various authors have commented on the nature
and significance of the contact between the Hatchery
Creek sequence and the underlying marine limestones,
but only some of these were based on actual field ©
investigations. Young (1969, p. 47) discussed the
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
upper limestone boundary, noting that the uppermost
Unit 6 of his ‘Upper Reef Formation’ was generally
poorly exposed because of high clay content, and
was covered by scree from the much more prominent
overlying “Hatchery Creek Conglomerate’ (now Wee
Jasper Formation). Where Unit 6 had continuous
exposure on the western shore of Lake Burrinjuck,
north from about GR491 243 around to the mouth
of Hatchery Creek, the beds were highly sheared in
the vicinity of the fold axis. The same applies at the
southern fold closure in the vicinity of the Long Plain
Fault south of Wee Jasper, obscuring sedimentary
changes at the boundary.
Young (1969) noted there was no change of
strike across the boundary, and no limestone clasts
were observed in the basal conglomerate. However,
in four measured sections across this interval
there was a marked difference in thickness of the
uppermost Unit 6, from 80 m in the south at GY39
(GR520 136), 210 m at GY40 (GRS508 183), 140 m
_at GY43 (GR499 210), and 110 m at GY44 (GR494
230). This thickness variation was attributed to
slight warping (less than 1°) before deposition of the
conglomerate, indicating a disconformable contact.
Pedder et al. (1970, p. 210) independently provided
similar evidence for a disconformable contact, noting
that the “Hatchery Creek Conglomerate’ (Wee Jasper
Formation) on the eastern limb ‘rests more than
250 feet above the highest assemblage zone of the
Taemas Formation, whereas on the western limb it
may rest less than 100 feet above the Hexagonaria
smithi smithi Teilzone’. They also noted that ‘the
lithologies of the two formations belong to entirely
distinct megafacies’. Owen and Wyborn (1979,
M320) also favoured a disconformable contact on
the evidence of thickness variation in the uppermost
unit of the Taemas Limestone, but suggested, from
the age evidence of the overlying fish assemblage
(subsequently published by Young and Gorter 1981),
that a ‘disconformity — if present — represents a short
time duration’.
Subsequent to these field investigations a new
track was cut around the western shore of the lake
at the northern end of the Goodradigbee valley. This
gave much improved exposure of this contact in
the vicinity of GR488 252, an important fossil fish
locality in the limestone (Fig. 2a). Here, Campbell
and Barwick (1999) measured a section through the
contact, the uppermost beds of the Taemas Limestone
comprising about 110 m of thin-bedded limestones
and shales ‘interpreted as an intertidal zone carbonate
deposit consistent with the fact that the overlying unit
is the fresh water Hatchery Creek Formation’ (p. 125).
Lindley (2002, fig. 4) presented a revised version
Proc. Linn. Soc. N.S.W., 131, 2010
of this section, with the uppermost unit beneath the
conglomerate assigned to Unit 6 of the ‘Upper Reef
Formation’ of Young (1969), and Campbell et al.
(2009, p. 62) noted that the top of carbonate sequence
with shallow marine algal mats was ‘transitional
into the overlying fresh water Hatchery Creek
Formation’.
Although uncertainty about this boundary was
indicated in stratigraphic sections of Basden et al.
(2000, fig. 2) and Young and Turner (2000, fig. 3B),
the new evidence just summarised is accepted as
indicating a conformable contact at the base of the
Hatchery Creek Group. The thickness variations
in the uppermost limestone units noted above must
therefore be interpreted as a depositional feature. This
complies with the original opinion of Edgell (1949,
p. 10) that interbedded lithologies at the contact
indicated continuous deposition.
The upper contact of the Wee Jasper Formation
(and base of the new Corradigbee Formation as
defined below) is at the top of Cycle D’ of Hunt
(2005). This is the highest cycle observed with
conglomerate/coarse pebbly sandstone forming the
basal unit, all higher cycles having sandstone at the
base (the rare thin conglomerates described below
for the Corradigbee Formation were within a cycle,
not at the base). It is noted that coarse beds persist to
the top of the Wee Jasper Formation in the vicinity of
localities 062 and 068 (Fig. 2a), but farther south the
equivalent beds seem less coarse, the contact being
less clearly defined, and recognised by a change in
colour rather than grainsize (discussed below).
Subdivision
The general outcrop of the Wee Jasper Formation
is indicated in Figure 1b, and a refined version of
Young’s (1969) subdivision into four units is detailed
in Figure 2. As noted above, the coarser basal unit
of each cycle (normally about 30-40 m thick), can
generally be traced with confidence on air photos,
although individual beds may pinch out along strike.
For example a prominent ridge just west of the Main
Ridge Trail at GR495 190 (Fig. 2a) is the next resistant
set of beds above the base of Unit 3, it forms the main
ridge for about 1 km along the track to the south, but
is less clearly differentiated in the Cave Creek type
section (Unit 3a, Fig. 2a). To the north it is traceable
to a similar prominent ridge immediately east of the
track at GR492 199, and it also crosses the track at the
Windy Top Trail turnoff. It forms prominent outcrops
immediately west of the track between GR490 208
and 4895 213, before it is crossed by the track again
at about GR488 219, where it is less distinct. This
is a distance of about 3 km along strike for what
12
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
Table 1. Faunal list for the Hatchery Creek fish assemblage (updated from Young and Gorter 1981).
Agnatha
Thelodontida
1. Turinia sp. cf. T. hutkensis Blieck & Goujet (Young & Gorter 1981)
Gnathostomata
Acanthodii
2. climatiid gen. et sp. indet.
?diplacanthiform gen. et sp. indet.
3
4. Tareyacanthus sp. cf. T. magnificus Valiukevicius (Burrow 2002)
5
Watsonacanthus? sp.
Osteichthyes (Sarcopterygii)
6. Gyroptychius? [new genus] australis Young & Gorter, 1981
7. osteolepiform gen. et. sp. nov. 2 (Hunt 2008)
8. osteolepiform gen. et. sp. nov. 3 (Hunt 2008)
2.
?onychodontid indet.
Placodermi
Arthrodira
10. Denisonosteus weejasperensis Young & Gorter, 1981
11. cf. Denisonosteus sp. nov. (Hunt 2005)
12. coccosteomorph cf. Coccosteus (Hunt 2008)
13. ?arthrodire gen. et. sp. nov. Hunt and Young, in press.
14. Arthrodira incertae sedis
Antiarcha
15. Sherbonaspis hillsi Young & Gorter, 1981
16. cf. Sherbonaspis sp. nov. (Hunt 2005)
17. Monarolepis verrucosa (Young & Gorter 1981) Young, 1988
is interpreted as a laterally discontinuous coarser
interval in the middle part of Unit 3.
The overlying recessive zone, representing the top
of Unit 3 at its boundary with the basal conglomerate
of Unit 4, is more persistent along strike, being
traceable over about 5 km back to the Cave Creek
Road type section. In the north it is crossed at a sharp
turn in the Main Ridge Trail at GR4855 221, it can be
followed south to GR4893 2015 (Windy Top Trail),
GR490 1955 (next valley south), GR4955 180 (east-
west section of Main Ridge Trail), and GR4955 1705
(Cave Creek Road type section).
Above this in the Cave Creek Road type
section, the coarse basal part for the overlying Unit
4 as mapped by Young (1969) corresponds to a sharp
bend in the Cave Creek road at GR495 170. Unit 4
is subdivided into 9 fining upward cycles (4a-j), the
upper parts of which correspond to the five ‘thin
zones of low weathering resistance’ mapped by
Young (1969). These are readily identified on recent
80
air photos in the valleys to the north and south of
the Windy Top Trail, designated here as type section
for the upper part of the Wee Jasper Formation. The
basal conglomerate/pebbly sandstone of Unit 4 (cycle
4a) is about 40-50 m thick, fining up into a poorly
outcropping interval of similar thickness, the latter
clearly visible on air photos as a continuous less
resistant zone from GR4845 224 south to the Windy
Top Trail type section. Here it separates the basal
conglomerate of Unit 4 at GR489 2015, and the basal
coarse beds of the second cycle, encountered at the
first bend in the track (GR488 202). This is the lowest
of three similar fining upward cycles (4b-d) crossed
by the track before a sharp southerly bend at GR4935
202. Each cycle is estimated at about 70 m thick, with
the coarse resistant beds comprising more than half
the thickness (4b, c), or about half (4d). These three
units are well exposed in the next creek to the south,
between about GR485194 and 490 196.
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
On air photos (and ‘Google Earth’) the E-
W sections along the valleys of the three creeks to
the north of the Windy Top Trail clearly show the
alternating resistant and five recessive beds of Unit
4 as mapped by Young (1969). The undifferentiated
upper part of the ‘Hatchery Creek Conglomerate’
of Young (1969) approximates to the Corradigbee
Formation as defined below. The upper part of cycle
Ac is the lowest of the five ‘less resistant mudstones’
mapped by Young (1969), and can be traced to the
north at least as far as the vicinity of GR478 222.
The recessive upper part of cycle 4d thickens
along strike to the north of the Windy Top Trail, in
the vicinity of GR483 205. The overlying four cycles
(4e-h) in this valley (the first creek north of the track,
between GR490 205 and GR475 206) are seen as
narrow ridges separated by less resistant bands of
equal or greater width. Most can be traced farther north
to the valley section of the creek between GR476 216
and GR487 216, where the resistant bands are thinner
_and recessive bands correspondingly thicker. The base
of cycle 4e is traceable to the south to cross the Windy
Top Trail immediately west of the sharp bend at GR483
200. Where the northern creek turns to the north-west
at GR476 216 the creek has eroded along the upper
recessive bed mapped by Young (1969). This is the
upper part of cycle 4f, traceable back to GR481 2005
on the Windy Top Trail. The basal coarse bed of cycle
4g is the lowest of three apparently thicker fining-
upward cycles (4g,h,j) along the Windy Top Trail,
their finer upper parts forming gullies immediately
to the south. However further south between about
GR475 194 and GR482 194 these beds are more
differentiated, and the less weathered outcrop along
the track may be due to relatively recent exposure by
removal of the overlying basalt. The uppermost of
these units (4j) passes beneath the remnant basalt cap
of Windy Top (Fig. 4b).
The correspondence between the uppermost
cycle 4j in the Windy Top Trail type section, and
Cycle C’ of the Corradigbee Formation as mapped in
the area farther south by Hunt (2005), is indicated in
Figure 2b. Cycle C’ is the lowest horizon in which
fish remains were found to the south, and in the gully
just south of the locked gate at Windy Top some
arthrodire fish fragments (ANU V2270) were found
at about GR476 200 by G. Young and A. Warren
in 1986, the equivalent lowest fish horizon in this
section. The interpreted correspondence between the
uppermost cycles identified are summarised in Fig. 2.
Figure 4b shows a view from the south towards Windy
Top, outlining the constituent units representing
uppermost cycles of the Wee Jasper Formation, and
the lowermost cycles of the Corradigbee Formation.
Proc. Linn. Soc. N.S.W., 131, 2010
Lithologies and sedimentary structures
Owen and Wyborn (1979: M314-M320) noted
numerous fining-upward conglomeratic cycles in
their type section. These varied in thickness from | to
20 m, partly due to upper beds in many cycles being
truncated by erosion such that one conglomerate
rested directly on the conglomerate of the preceding
cycle. A complete cycle was described in terms of
three lithologies. At the base they described a reddish
brown conglomerate, showing scoured contact
with the top of the preceding cycle, and including
subrounded to rounded pebbles and cobbles of
quartzite, quartz, chert, rhyolite and minor granitic
rock, with clay clasts and pellets. This was overlain
by reddish purple sandstone, usually thin-bedded and
flat-bedded, with local foreset cross-bedding (at about
20°). At the top of each cycle an upper red siltstone/
mudstone was described, with round whitish mottles,
containing root casts which bifurcate downwards,
extensively bioturbated in the upper part with bedding
sometimes completely destroyed, colour bleached
around numerous root casts; and rare wood tissue.
These cycles in turn make up the larger fining-
upward units mapped by Young (1969). The lowest
Unit 1 was described as 1-2 m thick conglomerates
interbedded with coarse lithic arenites for the lower
70 m, fining upwards into interbedded yellow
sandstones and red siltstones and mudstones. Unit 2
(thickness ~200 m) and Unit 3 (thickness ~400 m) are
similar fining upwards units, the basal conglomerate
of the latter exhibiting large scour and fill structures
at GY52 (GR499 1935), large scale cross-bedding
was recorded in overlying sandstones, and mudcrack
polygons in the upper part of Unit 3. Unit 4 (~700 m)
is generally finer grained, comprising more resistant
intervals 40-100 m thick separated by at least nine
thin zones of less resistant material summarised
in Figure 2b. In outcrop the more resistant strata
are pebbly sandstones up to 3 m thick interbedded
with red mudstone of similar thickness, although
considerable variation was observed (Young 1969, p.
50). The thin less resistant intervals, where examined
at two localities (GY50, 51, GR484 208, 4795 209),
are very distinct zones of no outcrop and sparse
vegetation about 10 m across, forming well defined
saddles on the crest of each ridge, with poor soil of
coarse red mudstone gravel presumably derived from
a friable red mudstone.
CORRADIGBEE FORMATION (NEW
FORMATION)
The change in lithology at about 1500 m recorded
81
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
by Owen and Wyborn (1979) was described as follows:
the conglomerate portion of each cycle becomes less
important, contains smaller pebbles, and in places is
absent, and the sequence is dominated by fine buff
sandstone and red siltstone with root casts. This finer
upper part approximates to the upper formation of
Edgell’s (1949) stratigraphic subdivision, and to the
new formation defined here, named after the property
(Corradigbee; GR64699 61166 on the Wee Jasper
1:25 000 topographic map 8627-4N, 2nd edition) that
encompasses much of its outcrop. Previous studies
referred to this unit as the ‘upper Hatchery Creek
Formation’ (Young and Gorter 1981; Francis 2003;
Hunt 2005), or ‘upper beds of the Hatchery Creek
Conglomerate’ (Owen and Wyborn 1979).
Detailed mapping in the study area of Hunt
(2005) revealed at least 18 sedimentary cycles in
this finer upper part, of which 15 are assigned to the
Corradigbee Formation. The base of its type section
(Figs. 1b, 3a) is at locality 063 (GR47598 17285),
and the top is at locality 082 (GR46644 18456). The
231.5 m section was measured in three parts, and the
composite section 1s given in Figure 3b.
Lower and upper contacts
The boundary between the Wee Jasper Formation
and the overlying Corradigbee Formation is defined at
the base of the fourth lowest cycle (Cycle C’). Cycles
D’ - F’ of Hunt (2005) correspond to the upper cycles
of the Wee Jasper Formation as described above (Fig.
2). The base of Cycle C’ is a fine sandstone, which is a
marked sediment change from the basal conglomerates
or coarse pebbly sandstones of all lower cycles. This
lithological change was observed in the northern part
of the field area at locality 068 (GR47793 18228),
extending to the north in the gullies immediately
south of the Windy Top type section. However, in the
southern part of the mapped area of Hunt (2005) the
underlying Wee Jasper Formation appears generally
less coarse than in the north, although these upper
beds were not mapped in detail. Along the access track
into Corradigbee homestead south of Goodradigbee
Hill (Fig. 3a) yellow sandstones predominate, and
conspicuous conglomerate or coarse sandstone strata
were not seen. The first conglomerates observed were
farther to the east (lower in the sequence) along the
main road (under the transmission line) in the vicinity
of GR475 155. In the vicinity of locality 063 (base
of the Corradigbee Formation type section), the
formation boundary was identified as a consistent
colour change, the underlying sediment (assigned to
the Wee Jasper Formation), including coarse grained
sandy-mudstone (containing root casts, bioturbation),
with a general very light yellowish brown colour.
82
In contrast, the overlying interbedded red and grey
mudstones containing fossil fish and plant material
(assigned to the Corradigbee Formation) is generally
much darker in colour. As a general impression the
grey mudstones seem to become darker in cycles
towards the middle part of the forn, \‘ion.
The uppermost horizons of the Corradigbee
Formation (K—M; see Fig. 4a) are exposed at
localities only in the core of the syncline, and only in
the southern part of the study area where erosion has
been impeded by the basalt cover. Another section
was measured on the western limb of the syncline
to include these upper cycles (Section 2, Fig. 3b).
The uppermost cycle M is inferred from a basal
sandstone overlain by about 2 m of mudstone before
cover by basalt scree. Thus the estimated thickness
of the Corradigbee Formation (260 m) is a minimum
estimate, because erosion before the basalt was
deposited is unknown.
Subdivision
Owen and Wyborn (1979) recorded at least three
grey sandstone — mudstone cycles in the upper fine-
grained part of the sequence, said to be less than 30
m thick and of limited lateral extent, each comprising
several sedimentary cycles. With more detailed
mapping, 15 sedimentary cycles are now identified
in the Corradigbee Formation, labelled from the base
to the top C’ to M (Fig. 3b), the original 1981 fossil
fish locality (WT) being in the third cycle from the
defined base of the formation (Cycle A). These cycles
are interpreted as cyclothems (i.e. an asymmetrical
repetition of sedimentary layers; Weller 1960). They
were first identified on air photographs by their basal
sandstones, which had a thickness greater than 20
cm. Two part sections were measured (sections la,
2, Fig. 3b), and compared with the type section to
demonstrate a similar sequence of cycles on both
sides of the syncline axis.
Cycle thickness varies, many being 12-15 m thick,
with an increase in thickness in the middle part of the
formation (Fig. 3b). This indicates either variation in
the period of time represented by each cycle, or more
likely variation in sediment supply, with the thicker
upper cycles reflecting increasing fine over coarse
material. These cycles indicate a repetitive sequence
of climatic or depositional conditions over the area,
presumably representing considerably longer time
intervals than annual cycles.
Lithologies and sedimentary structures
Owen and Wyborn (1979) described each fining
upwards cycle in terms of three lithologies: i) thin ©
basal medium grey coarse sandstone which contained
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
a Ib 260
< find M
basalt
scree
L 250
TYPE SECTION
200
Type Section 3) F
[069-075] Type Section
[063-067]
a
th
m
=
: 78 :
Corradigbee er
Goodrawigbee
Hil
=
ee Tei = Fossil Fish
erred a& Fossil Plants
T= > 065 © Carbonate nodules
TT Root Casts
i i = = = — gavel roads , 20
ae Se) es ight
—/ 063
Figure 3. a. Locality map for the study area of the Corradigbee Formation (base map Wee Jasper 1:25000
topographic map 8627-4N [second edition]). Previous fossil localities (JF, WT) and measured sections
indicated. For locality details see Appendix. b. Three measured sections through the Corradigbee For-
mation and suggested correlations. Locality numbers shown on the right of each section.
Proc. Linn. Soe. N.S:W.,13 1, 2010 83
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
Figure 4. a. View to the southwest from near the original fish locality at Windy Top, showing main
cycles of the Corradigbee Formation and position of measured sections. b. View to the north showing
the original fish locality (WT) to the west of the basalt cap at Windy Top, in the lower part of Cycle A.
The second fossil locality (JF, lower left) is in the upper part of the same cycle. Upper beds of the Wee
Jasper Formation (WJF) in the Windy Top type section shown to the right of the figure.
small subangular to subrounded pebbles; 11) thin fine
to medium-grained sandstone also including small
pebbles, and fish and plant fossils in one of the cycles
showing little evidence of abrasion, with fish plates
apparently not parallel to bedding, indicating that
the sandstone formed as a single bed; 111) an upper
dark grey to black massive mudstone up to 2 m thick,
containing vascular plant remains, rare fish remains
at the base, and grey-white limestone nodules in the
upper part, some containing microscopic fish remains,
and with mud cracks on upper bedding surfaces.
In the present study, lithologies can be described
in more detail for Cycle G of Section 2 as a typical
cycle (Fig. 3b). The base at locality 14 is a fine
84
sandstone (grain size <0.3 mm) approximately 3m
thick. Above the sandstone six mudstone/siltstone
units were identified by variation in colour. The first
3 m thick unit is a grey mudstone containing small
carbonate nodules (up to 5 cm diameter), in which no
fossils were found. This is overlain by another grey
mudstone about 7 m thick, containing both fossil fish
fragments and calcareous nodules. Above this is a 3
m orange mudstone layer, overlain by 1.5 m of dark
red mudstone, both lacking fossil material, followed
by a 5 m thick light grey mudstone producing
osteolepid and arthrodire fish material at locality 17.
Above this, another grey mudstone layer about 4.5
m thick contains large plant material (stems up to 30 —
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
Syncline axis
500 m
———:=~Creeks
Oo Fossil sites
Top of Wee Jasper
A Goodradigbee Hill “DE Formation
N @ Basalt Quarry
jee Tertiary Basalt
® Corradigbee HS
Figure 5. a. Outcrop map for the study area of the Corradigbee Formation. Cycles represented as alter-
nating shaded and clear units to indicate outcrop pattern. b. Measured dips and strikes in relation to the
syncline axis identified in this study.
cm in length) at the base, with fossil fish material and
scattered plants above. The next sandstone layer marks
the start of cycle H in this section (but correlated only
approximately with an additional sandstone in cycle
H of the type section).
Conglomerate
Conglomerates are very rare in the Corradigbee
Formation. One thin (~ 8 cm) bed of pebbly red
Proc. Linn. Soc. N.S.W., 131, 2010
conglomerate was observed at locality 70 (near the
middle of Cycle C). This contained small quartz
pebbles, mudclasts, and mudballs generally less
than 10 mm diameter, with generally rounded quartz
pebbles and grains, although some of the smaller
grains (<0.5 mm) were subangular. No fish fossils
were observed in the conglomerate bed, but these
occur immediately below in the mudstones (e.g.
ANU V3171). Another thin (up to ~5 cm) bed of grey
85
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
conglomerate was seen in erosion gullies at localities
98 and 138 (both probably in Cycle F), with quartz
pebbles up to 20 mm diameter, and much fragmented
fossil fish material giving the grey colour.
Sandstone
The sandstone layers at the base of each cycle
vary in thickness, over 3 m thick in cycle F (Fig.
3b), but most are about 1.5 m thick. Sandstones
within the cycles vary in the size of the sand grains
but grain size is uniform within the bed itself. Grain
size ranged from around <0.2 mm in each layer of
sandstone, with some beds being finer than others.
None of the sandstones in the formation were noted to
be very coarse grained. Good exposures of the basal
sandstones observed at localities 11, 14, 64, and 154
showed no cross-bedding, scour marks, mudclasts
or other evidence of a river deposit. At only one
locality (108, Cycle D) was some cross-bedding
observed. Fish material found in the sandstones was
disarticulated and fragmented (identified at only three
localities 134, 138, JF).
Siltstone/Mudstone
Siltstones and mudstones in the formation vary
in colour, from predominantly grey-black, to less
common orange, red, dark purple and light grey
lithologies. These colours are identified as primary
on the evidence that the colour terminated with the
bedding plane. In general, the red-purple colour
phases are assumed to have formed in well-drained
conditions, and the grey-black mudstones to indicate
poorly drained swampy conditions.
Sedimentary structures
In the mudstones of the Corradigbee Formation
calcareous nodules (up to 5 cm diameter) are abundant
at many levels (common at localities 62, 97, 109, 128,
137 and 158, but noted at many other localities). They
occur in both the red-purple and grey-black colour
variations (largest examples were seen at locality
158, in Cycle B). In the Devonian Aztec Siltstone of
Antarctica, common calcareous nodules were taken
to indicate lengthy subaerial exposure (4,000-10,000
years) for pedogenic processes to operate (McPherson
1979). The same can be assumed here, except that
the nodules are equally common in the red-purple
and grey-black colour phases, the latter representing
poorly drained swampy conditions, which would
preclude pedogenesis. Cubic pyrite crystals were
identified near fossil locality 161 (ANU 46692),
consistent with the idea that the black mudstones
formed under stagnant, anaerobic conditions.
86
Although laminar bedding was reported by
Young and Gorter (1981) and Francis (2003) to
indicate lacustrine conditions, only one occurrence of
laminar bedding was observed in this study, in grey
green mudstones at locality 24. Ripple marks were
identified at localities 129, 130 and 131. Rather than
lake deposits, the sedimentary structures indicate
predominantly swampy conditions for the Corradigbee
Formation, the whole Hatchery Creek sequence being
interpreted as a humid alluvial fan.
Root casts were noted at various levels in the red
and dark purple mudstones (Fig. 3b), in these cases
indicating sub-aerial exposure and soil formation as
do associated calcareous nodules. Apart from rain
drop impressions at locality 80, no other dessication
structures or mud crack horizons were observed in
this study.
STRUCTURE
Young (1969) recorded measurements from the
western side of the Goodradigbee valley indicating a
fairly consistent dip in the limestones and overlying
Hatchery Creek sequence, averaging 40° west with
a strike of about 338°. A plot of bedding/axial plane
cleavage intersections indicated a fold axis plunging
20-30° to the NW (315°). The uppermost limestone
beds forming the contact with the northernmost
exposure of the Hatchery Creek Conglomerate along
the edge of Burrinjuck Dam (on the Yass 100K sheet)
swing round a northern synclinal closure which
limited data suggested plunged about 35° to the
southwest (250°).
Owen and Wyborn (1979) showed only one
anomalous easterly dip on the Brindabella 1:100 000
geological map for the upper part of the Hatchery
Creek Conglomerate, interpreting the entire sequence
as dipping to the west, the basis for their estimated
2.9 km total thickness. They suggested renewed
uplift in the source area to explain a return to coarse
conglomeratic cycles at the top of the sequence, but
this can now be discounted (see above).
Their published cross sections (on _ the
1979 geological map) show the Hatchery Creek
Conglomerate as a thick westerly-dipping section
across the middle part of its outcrop (section A-
B), and tightly folded in the southeastern extremity
of the outcrop, with a steep to overturned western
limb against the Long Plain Fault Zone (section E-
F). Wyborn (1977) attributed this to thrusting of
the rigid Goobarragandra Block over the Hatchery
Creek Conglomerate, and no fold axis was indicated °
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
on the geological map. However, Edgell (1949) and
Pedder et al. (1970) had previously shown a syncline
axis running to the northwest towards the central
part of our Corradigbee Formation. This structure,
named the Wee Jasper Syncline by Hood and Durney
(2002), runs through the area mapped in detail by
Hunt (2005). New dip and strike measurements were
recorded from 69 localities in and around the area
of detailed mapping (see Appendix), and on both
sides of the syncline axis, which was identified in the
mapped area running through locality 21 and under
the basalt cap of the central ridge (Fig. 1b), which is
somewhat further to the west than the extrapolated
position shown by Hood and Durney (2002, fig. 1).
On the western side of the axis only easterly dips were
measured, conforming to the one anomalous easterly
dip of Owen and Wyborn’s map, and in the same area
Edgell’s (1949) map shows 10° and 13° easterly dips.
However, all measured dips were on the eastern side
of the Western creek (representative measurements
shown on Fig. 5b). We assume that the westerly
dips previously shown on the Brindabella 1:100 000
geological sheet for the upper part of Corradigbee
Formation outcrop must have been based on cleavage
masking the bedding.
SUMMARY
Type sections are proposed for a new Corradigbee
Formation, representing the upper fine-grained part
of the Hatchery Creek sequence, comprising about
15 fining-upward cycles of sandstones, dark shales
and mudstones in which 50 new fossil sites have been
found.
The lower coarse-grained part of the Hatchery
Creek sequence is renamed the Wee Jasper
Formation, within a revised Hatchery Creek Group
(total thickness about 1760 m). Thickness of the Wee
Jasper Formation is estimated at about 1500 m, it is
subdivided into four main fining upward cycles, and
an additional type section is nominated for the upper
part of the formation.
The Hatchery Creek Group is conformable on
Lower Devonian limestones of the Murrumbidgee
Group, thickness variations in the upper tidal flat
deposits of the carbonate sequence being interpreted
as depositional features.
Sedimentary structures indicate predominantly
swampy rather than lacustrine conditions for the
upper Corradigbee Formation, the whole Hatchery
Creek sequence being interpreted as a humid alluvial
fan.
Proc. Linn. Soc. N.S.W., 131, 2010
The axis of a major syncline was identified, with
previously unrecognised repetition of the lower coarse
strata in the western part of the outcrop area resulting
in a considerable over-estimate of total thickness in
published literature. The relatively high topography
of the softer shales and mudstones in the core of the
syncline is a relatively transient topography resulting
from recently eroded Tertiary basalts.
ACKNOWLEDGMENTS
For permission to conduct fieldwork on _ their
properties, and providing access, we thank Ian and Helen
Cathles of Cookmundoon (Wee Jasper), and Chris Barber
and Neil Blasford (Corradigbee). Access to the area for the
1969 field mapping was facilitated by Dudley and Graham
Barber. J. Gorter, A. Haupt, M. Owen, and R.W. Brown are
thanked for early field assistance. For assistance in 2003-
2008 fieldwork we thank B. Opdyke, K.S.W. Campbell,
I. Cathles, J. Caton, J. Francis, C. Klootwiyk, R. Hunt and
L. Bean. Professor Ken Campbell provided guidance and
knowledge on numerous occasions, and with Dr Brad
Opdyke gave helpful comments reviewing an earlier
manuscript. Professor S. Edgell is thanked for providing
a copy of his 1949 thesis and excellent geological map.
Val Elder (ANU) assisted in specimen curation and R.E.
Barwick with illustration. For comments on structural
geology we thank M. Rickard, S. Cox and D. Hood, and
on stratigraphic geology we thank K. Crook, A. Felton, and
D. Strusz. This research was supported by ARC Discovery
Grant DP0558499, and is a contribution to IGCP Project 491.
Provision of facilities at ANU in the Frank Fenner Building,
College of Science, and D.A. Brown Building, Research
School of Earth Sciences, is gratefully acknowledged.
REFERENCES
Basden, A., Burrow, C.J., Hocking, M., Parkes,
R. and Young, G.C. (2000). Siluro-Devonian
microvertebrates from southeastern Australia.
Courier Forschungsinstitut Senckenberg 223, 201-
DN.
Burrow, C.J. (2002). Lower Devonian acanthodian faunas
and biostratigraphy of south-eastern Australia.
Memoirs of the Association of Australasian
Palaeontologists 27, 75-137.
Branagan, D.F. and Packham G.H. (2000). ‘Field Geology
of New South Wales.’ (Department of Mineral
Resources New South Wales, Sydney, Australia.)
Campbell, K.S.W. and Barwick R.E. (1999). A new
species of the Devonian lungfish Dipnorhynchus
from Wee Jasper, New South Wales. Records of the
Australian Museum 51, 123-140.
Campbell, K.S.W., Barwick, R.E. and Senden, T.J.
(2009). Evolution of dipnoans (lungfish) in the Early
Devonian of southeastern Australia. Al/cheringa 33,
59-78.
87
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
Cramsie, J.N., Pogson, D.J. and Baker, C.J. (1978).
‘Geology of the Yass 1:100 000 sheet 8628’.
(Geological Survey of New South Wale, Sydney).
Edgell, H.S. (1949). ‘The Geology of the Burrinjuck-Wee
Jasper District.” B.Sc Honours thesis, Science Dept,
University of Sydney (unpublished), 75 pp.
Francis, J. (2003). ‘Depositional environment,
palaeontology and taphonomy of the Hatchery Creek
Formation, NSW.’ B.Sc Honours thesis, Geology
Dept, ANU (unpublished), 59 pp.
Hood, D.I.A. and Durney, D.W. (2002). Sequence and
Kinematics of multiple deformation around Taemas
Bridge, Eastern Lachlan Fold Belt, New South
Wales. Australian Journal of Earth Sciences 49, 291-
309.
Hunt, J. (2005). ‘An examination of stratigraphy and
vertebrate fish fauna of the Middle Devonian age
from the Hatchery Creek Formation, Wee Jasper,
New South Wales, Australia.’ Dept Earth and Marine
Sciences, Australian National University, B.Sc
(honours) thesis (unpublished), 113 pp.
Hunt, J. (2008). ‘Revision of osteolepiform
sarcopterygians (lobe-finned fishes) from the
Middle Devonian Hatchery Creek fish assemblage,
Wee Jasper, Australia.’ Research School of Earth
Sciences, Australian National University, M.Sc thesis
(unpublished), 109 pp.
Hunt, J. and Young, G.C. (in press). A new placoderm fish
of uncertain affinity from the Early-Middle Devonian
Hatchery Creek succession at Wee Jasper, New South
Wales. A/cheringa 35 (in press).
Joplin, G.A., Noakes, L.C. and Perry W.J. (1953).
“Canberra, New South Wales, 4-mile geological
series map. Sheet SI/55-16, Ist edition.’ (Bureau of
Mineral Resources, Australia).
Lindley, I.D. (2002). Acanthodian, onychodontid and
osteolepidid fish from the middle-upper Taemas
Limestone (Early Devonian), Lake Burrinjuck, New
South Wales. Alcheringa 26, 103-126.
McPherson, J.G. (1979). Calcrete (caliche) palaeosols
in fluvial redbeds of the Aztec Siltstone (Upper
Devonian), southern Victoria Land, Antarctica.
Sedimentary Geology 22, 267-285.
Miller, H. (1841). ‘The Old Red Sandstone.’ First edition.
(Edinburgh).
Owen, M. and Wyborn, D. (1979). ‘Geology and
Geochemistry of the Tantangara and Brindabella
area’. Bureau of Mineral Resources, Geology and
Geophysics, Bulletin 204.
Packham, G.H. (1969). The Geology of New South Wales.
Journal of the Geological Society of Australia 16,
1-654.
Packham, G.H. (2003). Discussion and Reply, Sequence
and Kinematics of multiple deformation around
Taemas Bridge, Eastern Lachlan Fold Belt, New
South Wales. Australian Journal of Earth Sciences
50, 827-833.
88
Pedder, A.G.H. (1967). Devonian rocks of the
Murrumbidgee River area, New South Wales,
Australia. In “International Symposium on the
Devonian System’ (Ed. D.H. Oswald) volume 2,
143-46.
Pedder, A.G.H., Jackson, J.H. and Philip, G.M. (1970).
Lower Devonian biostratigraphy of the Wee Jasper
region, New South Wales. Journal of Paleontology
44, 206-51.
Van Houten, F.B. (1973). Origin of Red Beds: a review-
1916-1972. Annual Review of Earth & Planetary
Sciences 1, 39-61.
Weller, J. (1960). “Stratigraphic principles and practice.’
(Harper and Brothers, New York).
Wyborn, D. (1977). Discussion - The Jindabyne Thrust
and its tectonic, physiographic and petrographic
significance. Journal of the Geological Society of
Australia 24, 233-236.
Young, G.C. (1969). ‘Geology of the Burrinjuck-Wee
Jasper area, N.S.W.’ B.Sc Honours thesis, Geology
Department, ANU (unpublished), 115 pp., 21 pls.
Young, G.C. and Gorter, J.D. (1981). A new fish fauna of
Middle Devonian age from the Taemas/Wee Jasper
region of New South Wales. Bureau of Mineral
Resources Geology and Geophysics, Bulletin 209,
83-147.
Young, G.C. and Turner, S. (2000). Devonian
microvertebrates and marine-nonmarine
correlation in East Gondwana: Overview. Courier
Forschungsinstitut Senckenberg 223, 453-470.
Young, G.C., Burrow, C., Long, J.A., Turner, S. and Choo,
B. (2010). Devonian macrovertebrate assemblages
and biogeography of East Gondwana (Australasia,
Antarctica). Palaeoworld 19, 55-74.
Proc. Linn. Soc. N.S.W., 131, 2010
J.R. HUNT AND G.C. YOUNG
Abbreviations:
APPENDIX
f= fish, p = plants, n =nodules, a = arthropods, r= root casts.
b = bioturbation, g = gastropods, impr = rain drop impressions.
x = cross bedding, lam = laminar bedding
2005 Localities
I 55 H 645457 6117228
2 55 H 646000 6119045
3 55 H 645626 6116656
4 55 H 646151 6118053
5 55 H 646519 6119415
6 55 H 646535 6119428
7 55 H 646646 6119620
8 55 H 647280 6118881
9 55 H 647194 6118825
10 55 H 646906 6118728
11 55 H 646062 6118133
12 55 H 646093 6118127
13 55 H 646125 6118108
14 55 H 646121 6118105
15 55 H 646154 6118073
16 55 H 646149 6118087
yl 55 H 646170 6118087
18 55 H 646179 6118060
19 55 H 646542 6118600
20 55 H 646426 6118514
21 55 H 646418 6118061
22 55 H 646455 6117973
23 55 H 646600 6116644
24 55 H 647449 6116390
59 55 H 646709 6116402
60 55 H 646585 6115979
61 55 H 646570 6116005
62 55 H 646552 6116007
63 55 H 646655 6116134
64 55 H 646543 6115914
65 55 H 646450 6115880
66 55 H 646199 6116092
67 55 H 646261 6116124
68 55 H 646237 6116362
69 55 H 646517 6116499
70 55 H 647411 6117304
il 55 H 647306 6117767
72 55 H 647331 6117878
73 55 H 647310 6118258
Tes 55 H 647373 6118281
Proc. Linn. Soc. N.S.W., 131, 2010
nt O)
waAdo
(SI tex) ee le) 2 fool (esl tes) lel GQ leat [eal les! fest esl esl ) Ie! yal fae} [3e| jae) fae, GY EG) © 'aal 'sl
GRID REFERENCE Horizon Dip/Strike Fossils and
Structures
6°E/345° f
p
f
f
13°W/106° f.p
f
7 W/25° x
19°W/10°
5°E/345°
f
f.p.n
f
10E*/295°
10N/310°
p
10°N/285° __ f. lam
15°N/95° f
p
f
n
p
5°E/120°
p
10°E/120°
f
23°W/5° f
r
17° W/210°
6° W/138"
p
89
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
15 55 H 647458 6118317 D _18°W/135° rb
76 55 H 647754 6117945 B 15°W/130°
af, 55 H 647465 6116410 A 9°W/148"
78 55 H 647452 6116390 B 11°W/65°
79 55 H 646479 6117559 M f
80 55 H 646475 6117509 M 8°W/345° ff. impr
81 55 H 646432 6117481 L 10°E/185° f
82 55 H 646408 6117475 J f
83 55 H 646410 6117566 J \S*B/SF
84 55 H 646404 6117569 J -9°E/140 f
85 55 H 646309 6117780 Ib
86 55 H 646280 6117839 J 8°E/30 f
87 55 H 646238 6117838 L 8°E/30 f
88 55 H 646226 6117873 J f
89 55 H 646218 6117934 J -8°E/50° f
90 55 H 646224 6118003 1 ena f
91 55 H 646235 6118092 J -11°E/180 fia
92 55 H 646250 6118240 I f
93 55 H 646422 6118054 I 10°N/110°
94 55 H 646398 6118027 J 8°E/120°
95 55 H 646433 6117488 by
96 55 H 646342 6117551 I p
My) 55 H 646316 6117507 H f
98 55 H 646279 6117544 I 14°E/310° f
99 55 H 646207 6117646 G 5°E/310°
100 55 H 646489 6119386 H 3°W/325° f
101 55 H 646505 6119399 F f
102 55 H 646506 6119407 E f
103 55 H 646522 6119419 D 6°W/315° f
104 55 H 646544 6119454 E
105 55 H 646572 6119474 D 6°W/335° f
106 55 H 646509 6119660 D fp
107 55 H 646487 6119618 D 8°W/335°
108 55 H 646382 6119535 D 8°W/340° p
109 55 H 646368 6119522 D fin
110 55 H 646218 6119567 H 3°W/310°
111 55 H 646118 6119555 J
112 55 H 645876 6119708 H
113 55 H 645773 6119735 De RAS:
114 55 H 645727 6119736 D 3°E/325°
IS 55 H 645738 6119686 Dy = SBS:
116 55 H 645779 6119398 C
117 55 H 645788 6119395 C 3°E/335°
118 55 H 645824 6119392 H f
119 55 H 645914 6119367 J
120 55 H 646583 6118459 I f
121 55 H 646753 6115769 H p
122 55 H 646740 6115768 G 3°E/330°
123 55 H 646371 6115865 G 6°E/335°
124 55 H 646453 6115883 E 5°E/356° p
90 Proc. Linn. Soc. N.S.W., 131, 2010
125
126
U27
128
WY)
130
31
Iz
133
134
3s)
136
1397,
138
139
140
141
142
143
» 144
145
146
147
148
149
150
esi
152
153
154
ISS
156
157
158
SQ
J.R. HUNT AND G.C. YOUNG
55 H 646090 6115893
55 H 646011 6115953
55 H 645950 6116511
55 H 645483 6120425
55 H 646315 6121384
55 H 646380 6121374
55 H 646394 6121386
55 H 646141 6119126
55 H 646036 6119138
55 H 645968 6119128
55 H 646244 6117558
55 H 645697 6119296
55 H 645734 6119219
55 H 645965 6119015
55 H 645848 6118845
55 H 646434 6116715
55 H 646297 6117036
55 H 646318 6117284
55 H 645979 6118978
55 H 645936 6119145
55 H 645938 6119144
55 H 645947 6119148
55 H 645923 6119179
55 H 646075 6118902
55 H 646113 6118668
55 H 646115 6118654
55 H 646172 6118227
55 H 647697 6117736
55 H 647329 6117876
55 H 647197 6117955
55 H 647126 6118159
55 H 647090 6118284
55 H 647262 6118401
55 H 647375 6118448
55 H 647245 6116898
Appendix continued p. 92
Proc. Linn. Soc. N.S.W., 131, 2010
ie)
ap
ies) los lool I) to) js lee) lee) @) Jacl 'o5] Iq] Ioaf M5] [9 'a5) jae
E/3555
4°E/30°
6°E/350°
11°W/330°
20°W/345°
34°W/350°
lOME/S35%
TPYSO
© Als
6°E/310°
6°E/40°
SE Silss
me io2 0
18°W/350°
10°W/310°
10°W/345°
14°W/10°
6°W/35°
nN
Mh
STRATIGRAPHIC REVISION OF THE HATCHERY CREEK SEQUENCE
2007/08 Localities
160
161
062
063
064
065
066
067
068
069
070
071
072
073
074
075
076
077
078
079
080
081
082
92
GRID REFERENCE
55 H 645457 6117228
55 H 646597 6118044
55 H 647714 6118047
55 H 647598 6117285
55 H 647328 6117162
55 H 647280 6117140
55 H 647188 6117113
55 H 647793 6118228
55 H 647766 6118251
55 H 647173 6117114
55 H 647155 6117111
55 H 647114 6117413
55 H 646093 6117410
55 H 647073 6117404
55 H 647068 6117404
55 H 647031 6117389
55 H 647006 6118570
55 H 646982 6118563
55 H 646946 6118572
55 H 646905 6118579
55 H 646851 6118549
55 H 646644 6118456
Horizon Dip/Strike Fossils and
2s
2299-
VA; Cai iS JRE, fae) jae) jae) jac; jae, (@) 'ol les! lo) lee} () lee! =
Structures
6°E/345°
p-g
11°W/320° b
b
14°W/352°
n
f.p
p.n.b
14°W/320° if
6° W/340° n
14°W/40°
14°W/350°
if
Proc. Linn. Soc. N.S.W., 131, 2010
Reproductive Phenology of White Box (Eucalyptus albens
Benth.) in the Southern Portion of its Range: 1997 to 2007
W.S. Sempce! AND T.B. KOEN?
' Formerly Department of Environment and Climate Change, Orange, NSW 2800. Present address: 37 Popes
Rd, Junortoun, Victoria 3551 (b.semple@bigpond.net.au)
* Department of Environment, Climate Change and Water, PO Box 445, Cowra, NSW 2794.
Semple, W.S. and Koen, T.B. (2010). Reproductive phenology of white box (Eucalyptus albens Benth.)
in the southern portion of its range: 1997 to 2007. Proceedings of the Linnean Society of New South Wales
131, 93-110.
The abundance of reproductive structures (buds, flowers and capsules) in individual Eucalyptus
albens trees at four sites was monitored for up to 11 years. Average abundance values for a stand of trees
often masked individual differences, e.g. abundant budding (a surrogate for flowering) in consecutive years
was never recorded in a stand but 1t was common in individuals. On average, floral buds appeared in
November and flowers were produced between March and November the following year but some trees
produced buds as early as March, and in others flowering extended to January. Though summer-flowering
was uncommon in this study, some observations from the 1970-80s reported a flowering period of, for
example, January to June, suggesting that flowering is now later. Except for peak flowering years, e.g. at
three sites in 2006, when virtually all trees flowered, flowering was individualistic suggesting that previous
rainfall was not the sole driver. Correlations between bud abundance and previous rainfall suggested
that individual trees, or groups of trees, responded to different rainfall events. For example, budding in
some trees at all sites (particularly those in the two northern-most sites) was positively correlated with
winter rainfall three years previously whereas at the most southerly site, budding in many of the trees was
correlated with autumn rainfall four years previously. Such variability may be genetically determined and
have positive benefits for seedling recruitment in a variable climate such as Australia’s.
Manuscript received 18 February 2009, accepted for publication 17 February 2010
KEY WORDS: capsules, Eucalyptus albens, floral buds, flowers, rainfall, seedling recruitment, variability
“All around Sydney, and particularly in our bushland suburbs, the Angophora costata
(Sydney Red Gum) are in exceptionally heavy flower. So heavy that the white honey scented
blossoms weigh the branches down to give the trees an uncharacteristic domed shape. Why are they
busily preparing for such a profusion of seeds to drop this year? What do they know that we don t?”
Letter to the editor, Sydney Morning Herald, 27 November 2006
INTRODUCTION
Woodlands dominated or co-dominated by
white box (Eucalyptus albens) once extended almost
continuously from southern Queensland, along the
inland slopes of New South Wales (NSW) into north
central Victoria with outliers in the Snowy River
area, western Victoria and the Southern Flinders
Ranges of South Australia. The woodlands occur on
several soil types that, at least for those with a grassy
understorey, are relatively fertile and are now used
for wheat-growing (Beadle 1981). Consequently the
woodlands now occupy a lesser area than they once
did. Nevertheless E. al/bens trees are still relatively
common across their range and contribute to the
aesthetics of the roadsides and farmlands where
they occur. However, intact grassy woodlands, 1.e.
those with relatively undisturbed overstorey and
groundstorey, are rare and poorly conserved in the
formal reserve system (Prober 1996). They are listed
nationally as an endangered ecological community
REPRODUCTIVE PHENOLOGY OF WHITE BOX
under the Environment Protection and Biodiversity
Conservation Act 1999,
Natural recruitment of seedlings of E. albens
is uncommon, at least in the southern part of its
range, and has been attributed (Semple and Koen
1997, 2003) to the seedling’s inability to compete
with exotic species that are now dominant in many
groundstoreys of these woodlands. Exotic dominance
is probably due to enhanced soil fertility, particularly
nitrogen (Prober et al. 2002) and/or phosphorus
(Allcock 2002). Other potential limitations to
successful seedling recruitment include: reduced
seed quantity and quality produced by isolated trees
in cleared environments (Burrows 1995), the unlikely
coincidence of suitable rainfall for both germination
and survival, browsing of seedlings by wingless
grasshoppers and domestic and feral animals, minimal
seed reserves in the soil due to predation by ants and
ready germination of non-dormant seed following
rainfall events. A consequence of the last-mentioned
is a reliance on an aerial seedbank from which seed is
shed intermittently (Semple et al. 2007).
The amount and occurrence of seed fall is
primarily determined by a range of prior factors that
affect the production of buds and in turn, flowers
and fruits. In the case of eucalypts, the inflorescence
commences as a bud that differentiates into a cluster
of ‘bud initials’ (‘inflorescence buds’) that are
enclosed by a cap of fused bracts. After the cap is torn
and shed, buds develop through ‘pin’, ‘cylindrical’
and ‘plump’ stages until anthesis (Boland et al.
1980). Each bud consists of a basal hypanthium, in
which the ovary is wholly or partially embedded,
and the calyptera (operculum), which encloses
the stamens. In species of the Symphomyrtus sub-
genus, the operculum is double-layered and the outer
calyptra is shed early or, as in the case of E. albens,
fuses with the inner, which is shed at anthesis (Hill
1991). Following pollination (by insects, birds, small
mammals) and fertilization of ovules, seed and fruit
development commences. Fruits (capsules) expand,
change colour from ‘green’ to ‘brown’ and become
increasingly woody. Dehiscence is initiated by twig
death or the formation of an abscission layer that cuts
off the sap flow to the capsules. Fertilised ovules are
shed as seed and unfertilised ones (the majority) and
ovulodes as ‘chaff’.
In an earlier study of E. albens trees near Cowra,
NSW, Semple et al. (2007) reported that seed fall was
highly variable between trees as was the occurrence
and abundance of flowers. Moderately abundant
flowering occurred every second year on average
and appeared, at least in the period 1996 to 1999, to
be associated with above-average rainfall in winter
94
and spring the previous year. Whether biennial
flowering was usual or whether it was associated
solely with previous above-average rainfall could not
be determined from data that was limited to scattered
paddock trees at one site and only four years of
observations.
The study reported below formed a component
of a broader study investigating the role of various
factors (seedbed, rainfall, seed fall, etc.) in the seedling
recruitment of woodland eucalypts. It aimed to (a)
document the seasonality, frequency and abundance
of floral buds, flowers and capsules in individual trees
within stands that were distributed across the southern
range of E. albens; and (b) examine the relationship
between rainfall and the production of floral buds
over a longer period than was the case at Cowra.
METHODS
Site selection
The basic requirements were for stands containing
at least 12 trees of variable size, as indicated by
diameter at breast height (DBH), that were readily
(and safely) accessible. The latter was satisfied by
occurrences beside roads that were travelled regularly
in the course of normal business or recreation. Small
trees that were unlikely to flower were ignored but
these were only evident at one site (Molong). An
additional requirement was that stands were distributed
relatively evenly across the southern distribution of
the species, viz. from central western NSW to north-
eastern Victoria. There were no requirements with
respect to aspect, altitude or condition of the stand
though those with unhealthy, e.g. dieback-affected,
trees were avoided. Four sites, located to the north
and south of the earlier study site near Cowra, were
selected (Fig. 1). All stands were parts of ‘corridor
communities’ (e.g. Fig. 2) except at Molong where
the stand extended into the adjacent paddock. None
was located near a supplementary source of water,
such as a dam or watercourse, and spatially variable
run-on (with associated nutrients) from the roadside
or adjacent land appeared unlikely. An unintended
consequence of the selection procedure was that as
latitude increased, altitude and mean annual rainfall
generally decreased (Table 1).
Monitoring
Trees were observed with binoculars by the same
observer [WS] at regular intervals — ideally monthly
during bud formation and flowering (usually mid/
late autumn to late spring, when new floral buds also
become evident). At each observation the abundance
Proc. Linn. ‘Soc: N.S.W., 131, 2070
W.S. SEMPLE AND T.B. KOEN
33°0'S
Mid Western Hwy
150°0'E
Figure 1. Location of towns nearest the four E. albens sites in the present study and an earlier one near
Cowra.
of reproductive structures across the canopy of each
tree was assessed on a 6-point integer scale: 0 (none),
1 (one to very few), 2 (scattered or a few small
clumps), 3 (obvious and dispersed across most of the
canopy), 4 (very abundant), 5 (maximum possible).
Structures assessed were: pin buds, buds (‘cylindrical’
and ‘plump’ stages were not distinguished), flowers
(up to withering of anthers) and capsules (= all post-
flowering structures with no distinction made between
fruits at different stages of maturity). Initial attempts
Proc. Linn. Soc. N.S.W., 131, 2010
at assessing ‘inflorescence buds’ were abandoned
as they could not be distinguished reliably from the
vegetative buds that were produced each autumn and
spring with the latter period often coinciding with the
presence of inflorescence buds. Observations were
less frequent over summer and also during periods
when bud production was nil or minimal (and hence,
flowering was unlikely to occur). Inevitably over a
monitoring period of up to 10 years, there were periods
when bud and/or flower activity were missed.
95
REPRODUCTIVE PHENOLOGY OF WHITE BOX
Figure 2. A typical roadside stand of E. albens. The monitored stand at Yerong Creek in November 2006
[photo 245/6].
Table 1. Brief details on the monitored roadside stands of Eucalyptus albens listed in order from north
to south.
Mean
Locality and ee DBH "(m): mean Altitude annual Period of regular
Stand name : at start ; or Wego
latitude A and range (ma.s.l.) rainfall monitoring
(end “)
(mm)
6 km SW of
Molone nee , 13) C12) eee OlG 2) (ONINESTSS) 600 700 Mar. 2000 — Nov. 2006
Rest area, 7.2 km
Young Dy evening 12(11) —-0.67 (0.41-1.15) 550 650 July 1997 — Nov. 2006
34°17'12"'8
3.6 km S of
Meronp@reseuy eeepc 19(18) 0.52 (0.14-0.99) 230 530 ‘Jan. 1997 — Nov. 2006
35° 25'00"S
Rest area, 6kmS$S
Springhurst See a 19(18) 0.54 (0.18- 2.08) 180 610 Dec. 1996 — Nov. 2006
A Tree decline was due to deliberate removal associated with roadworks (Molong and Springhurst), ringbarking (Young
shortly after observations commenced) and tree fall (Yerong Creek).
B Diameters of any multi-trunked trees have been summed.
C All stands were revisited in early 2007 to assess the size of the 2007 bud crop though Molong observations were ig-
nored because of the confounding effects of a wildfire in November 2006.
96 Proc. Linn. Soc. N.S.W., 131, 2010
W.S. SEMPLE AND T.B. KOEN
Regular monitoring ceased in November 2006
though the bud crop for 2007 was assessed on number
of occasions at all sites except at Molong where most
of the trees were severely burnt in November 2006
[though monitoring at this site was maintained so as
to document the effects of fire on the trees and the
groundstorey (see Semple and Koen 2008)].
Data analysis and presentation
Data for all types of floral bud have been
amalgamated for presentation purposes. Where
trees were not observed as frequently as desired
(i.e. missing monthly observations), the abundance
of reproductive structures has been interpolated
when little change was known to have occurred.
However, where new structures appeared between
these extended observation periods, the periods of
unobserved activity have been shown as ‘missing
data’ on graphs of abundance of structures.
Averaging the abundance ratings of flowers
_ across all trees at a site at each time of observation
was misleading because individual trees flowered
over varying periods of time (or failed to flower at
all) and times of maximum flower abundance in
individual trees did not always coincide. Hence,
average values across the flowering season implied
lower abundance than was the case. Conversely,
floral buds usually developed synchronously in trees;
and averages of maximum values prior to flowering
provided an indicator of potential flowering in a stand
in any one season. Bud abundance has generally been
used as a surrogate for overall flower abundance in
the analyses presented here.
The suggestion that larger/older eucalypts
flower more frequently and heavily than smaller ones
(various authors cited by House 1997) was examined
via correlations between DBH and the frequency of
abundant budding (abundance rating >3) of trees
at each site. Two sets of DBH values were used —
averaged and summed DBHs for multi-trunked trees.
Associations between rainfall and bud abundance
were examined for each site and for each tree. The
interpolated monthly rainfall (Jeffrey et al. 2001) at
each site was summed in various periods: calendar
year, warm (September to February of following year)
and cool (March to August) season, and actual season
(autumn, winter, etc) for each year of data, 1986 to
2006. Linear correlations were calculated between
each of these rainfall periods and the maximum bud
abundance (usually in summer each year) for (a) each
site (mean values), and (b) for each tree.
Proc. Linn. Soc. N.S.W., 131, 2010
RESULTS
Abundance of buds and flowers in stands
Average abundance ratings for floral buds and
flowers over time at the four sites are presented in Fig.
3. Low abundance ratings (<3) generally indicated
very low numbers of structures and can largely be
ignored — apart from cases of flowering at low levels
over an extended period. The occurrence of abundant
budding (mean rating of >3) was uncommon at most
sites: three in seven years at Molong, three in nine
years (ignoring incomplete data for 1997) at Young,
two in 10 years at Yerong Creek and Springhurst.
Between these abundant budding years, at least some
of the trees produced buds and flowers, sometimes at
very low levels, except at Springhurst in 1997, 1998,
2000 and 2001 when no buds or flowers were observed
(though very low level budding and flowering may
have been missed).
Periods of abundant budding tended to occur
every second or third year but were less frequent at
Springhurst. Some stands budded abundantly in the
same years (e.g. 2001 and 2006) but the sequence of
budding in the two southerly stands, particularly at
Springhurst, was usually different from those in the
north. Years of high average bud abundance were
followed by at least one year of low abundance.
Abundant budding levels in each stand were positively
associated with the proportion of trees producing
abundant buds in that year (compare Figs. 3 and 4).
Times of bud formation and flowering in stands
Pin buds were usually evident between October
and December. Buds were at a maximum by early
summer and abundance ratings rarely declined prior
to the commencement of flowering.
During peak flowering periods when most trees
flowered abundantly, flowering in some trees was
usually evident in March (though as early as February
in some trees at Young in 2003; Fig. 5a) with the latest
commencing in June or July. Flowering was usually
complete in all trees by October or November. Some
trees flowered for a long period between March and
November but most trees flowered for only a few
months. In non-peak flowering years when only some
trees flowered, some trees, usually those with very
few buds, did not commence flowering until August
or September.
Some of the Molong trees did not follow these
trends. For example, the main flowering period for
tree M194 in 2003 was from November to January
2004. Some trees produced pin buds very early in the
season: two trees (M181 and M192) during March/
August 2000 and one tree (M181 again) in May 2002;
97
REPRODUCTIVE PHENOLOGY OF WHITE BOX
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
°° Mean bud abundance rating / tree
>» Mean flower abundance rating / tree
4 | Springhurst | 4
"4997 1998 1999 2000. 2001 2002 2003. 1998 1999 2000 2001 2002 2003 2004 2004 2005 2006. 2006
Figure 3. Mean abundance ratings (0-5) for floral buds (0) and flowers (A) over varying periods of times
at four stands of E. albens. Sites are presented in order from north to south. Periods of missing data have
generally been smoothed over except when bud initiation, or a major flowering event (i.e. Yerong Creek
in 1998), were missed.
but these buds matured slowly and were eventually Unusually, a small number of buds that became
indistinguishable from buds produced at the normal evident in October/November at Molong produced
time (~November). Small quantities of early pin buds flowers in November/January. This occurred at trees
were also produced by a few other trees at Molong,and M177, M192 and M181 in 2003, 2004 and 2005 |
one at Young, but they apparently failed to develop. _ respectively (Fig. 5b)
98 Proc. Linn. Soc. N.S.W., 131, 2010
W.S. SEMPLE AND T.B. KOEN
1997 1998 1999 2000 2001
100
| Molong (12).
| Yerong Creek (18)
Proportion (%) of trees with abundant buds
pepiinghvisnttey:
1997 1/993
1999 2000 2001
2002 2003 2004 2005 2006 2007
2002 2003 2004 2005 2006 2007
Figure 4. Proportion (%) of trees in each stand that produced abundant (rating > 3) floral buds in any one
year. * = nil or incomplete data. Numbers of trees monitored for the full period at each site are shown in
parentheses. Bud abundance in 2007 was determined from a few strategically-timed observations.
Budding and flowering of individual trees within
stands
Frequency, abundance and duration of flowering
varied between trees at all sites, particularly in years
when flowers were not abundant. Space prohibits
the presentation of all data. Young and Molong are
presented as examples in Figs. 5a and 5b. During
the ‘big’ budding/flowering years at Molong (2001,
Proc. Linn. Soc. N.S.W., 131, 2010
2004, 2006 and to a lesser extent 2003), Young (2001,
2003 and 2006), Yerong Creek (1998 — presumably
as the main flowering period was missed, 2001 and
to a lesser extent 2004) and Springhurst (1999 and
2006), all trees flowered — except for one or two trees
at Springhurst in 1999 and Yerong Creek in 2004
— though with varying levels of intensity.
99
REPRODUCTIVE PHENOLOGY OF WHITE BOX
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
5
4
3
2
1
» 5
4
3
2
1
5 0
4
3
2
1
2 5
4
3
2
1
0
(4D) 5
Of 4
35 4
=< 1© 2
2) 0 ie
=e 5
—_ a 4
= 3
® » 1
(Oe 1S :
Se
oe 4
2 pe
= ra
2 | 1
io aoe 5
Y6 (0.59) 4
3
2
1
5 | eet | a N 0
s ¥ oo ¥10 (0.52) P
2
1 2
4
3
2
1
(0)
OpmP-NWHW
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Figure 5a. Floral bud (0) and flower (A) abundance ratings (0-5) for eleven E. albens trees on a roadside —
near Young: July 1997 to November 2006. Tree identification numbers are preceded by the letter Y, and
have DBH (m) in parentheses.
100 Proc. Linn.. Soc. N.SEW.,.13 1, 2010
W.S. SEMPLE AND T.B. KOEN
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
5 CTT
44 M177 (0.97) ;
3
2
1
0
5
4
3
2
1
& 0
4
3
2
1
0
5
4
3
2
1
cD) 5 °
og 5
— = 3
~ Zz 2
Dug o
Ss ais 5
a 4
o 3
O wow 2
o's 1
Sm 5 (0)
oe -
mos 4
Sai oe
= 2
S 4 sn
2
4
3
2
1
2 0
4
3
2
1
5
4
3
2
1
é 0
4
3
2
1
: 5
4
3
2
1
0
OrFNWAMN
1997 1998 -1999 2000 2001 2002 2003 2004 2005 2006
Figure 5b. Floral bud (0) and flower (A) abundance ratings (0-5) for thirteen E. albens trees on a road-
side near Molong: March 2000 to November 2006. Tree identification numbers are preceded by the letter
M, and have DBH (m) in parentheses. * = no data.
Proc: linn: Soc. N.S:W.,.1/31), 2010 101
REPRODUCTIVE PHENOLOGY OF WHITE BOX
Abundantbudding (mean rating >3)inconsecutive
years across a stand was rare (Fig. 3) but it was often
recorded in individual trees. At Young, three trees
(Y2, Y5, Y7) budded abundantly in consecutive years
on one occasion, and another (Y 10) on two occasions
(Fig. 5a). Abundant budding in consecutive years
was less frequent in trees at Springhurst (two trees on
one occasion each) but considerably higher at Yerong
Creek: eight trees on one occasion and four trees on
two occasions but in the case of two of the latter trees,
the second occasion extended over four years, 2001
to 2004. Despite the shorter period of observation
at Molong, four trees (M177, M309, M194, M255)
produced abundant buds in consecutive years on one
occasion and five trees (M181, M192, M238, M250,
M223) on two occasions — though in some cases buds
declined prior to flowering, e.g. at M255 in 2003 (Fig.
5b).
Some trees budded abundantly more often than
other trees at all sites (Fig. 6). This was particularly
evident at Molong where five (41%) trees budded
abundantly in five of the seven years observed. At
the other extreme, six trees at Springhurst produced
abundant buds in only one of the 11 years observed.
Larger trees tended to produce abundant buds more
frequently than smaller ones, at least for the range of
DBHs shown in Table 1, but the overall association
was low, ranging from r = 0.23 at Young to r = 0.70
at Molong.
Production and decline of capsules
The abundance of capsules in individual trees
over time reflected the varying flowering patterns, and
minor flowering events (bud abundance <2) generally
had an imperceptible effect on the crop of capsules.
Though peak flowering events (Fig. 3) were
important in replenishing the capsule crop in stands
(Fig. 7), even minor flowering events (mean bud
abundance <2) played a role because some trees
flowered abundantly during these periods. Though the
crop consisted mainly of immature capsules following
each peak flowering, for much of the time crops of
different ages were present in the canopies — except at
Springhurst where flowering was infrequent. For most
of the time at this site, average capsule abundance
was low (<2) and any fruits present were likely to
have been over-mature, i.e. dehisced.
Relationship between the occurrence of budding
and preceding rainfall
Linear correlations were examined primarily for
significant correlations between bud abundance and
recent (< 5 years previously) rainfall that the site (i.e.
mean values) shared with many of the individual
102
trees. A subset of the rainfall data, cool-season and
warm-season, is presented in Fig. 8.
Mean maximum bud abundance at Molong was
significantly correlated (7 = 0.81) with winter rainfall
three years previously (Fig. 9a) and warm-season
rainfall five years previously (7 = 0.82); and negatively
correlated with cool-season (r = —0.76) and/or winter
(r = —0.78, Fig. 9b) rainfall four years previously.
Only three trees exhibited all correlations but most
showed one or two. Bud abundance at four trees
(M181, M192, M238, M250) was not significantly
correlated with recent rainfall.
At Young, mean maximum bud abundance was
also significantly correlated (r = 0.69) with winter
rainfall three years previously (Fig. 9c) and negatively
with winter rainfall four years previously (r = —0.64,
Fig. 9d) but also with summer rainfall one year
previously (r = 0.72). None of the individual trees
showed all three correlations. Bud abundance for the
first five trees in Fig. Sa was correlated with winter
rainfall three years previously and summer rainfall
one year previously. Figure 5’s last three trees, which
tended to produce abundant buds in most years, were
not consistently associated with these lagged rainfall
series but bud abundance at two of them (and also
Y2) was significantly negatively correlated with
winter rainfall four years previously.
Mean maximum bud abundance at Yerong Creek
was significantly correlated with spring (r = 0.62)
and/or warm season (7 = 0.64) rainfall three years
previously. Budding at seven of the 18 trees with a
complete set of data showed a similar pattern. Unlike
Molong and Young, the positive correlation with
winter rainfall three years previously and the negative
correlation with winter rainfall four years previously
were evident at only one or other of four trees, and
across all trees these correlations were weak (Figs.
9e and 9f).
At Springhurst, mean bud abundance was
significantly negatively correlated with rainfall two
years previously: calendar year (r = —0.72) and cool-
season (r = —0.66). One or both of these correlations
were evident for 13 of the 18 trees with a complete
data set but budding at nine trees was also significantly
positively correlated (7 values ranging from 0.60
to 0.74) with autumn rainfall four years previously.
Correlations with winter rainfall three and four years
previously were weak (Figs. 9g and 9h).
Across all 59 trees, bud abundance at 24 was
significantly positively correlated with winter rainfall
three years previously. Such trees were present at all
sites, particularly at Molong and Young. At Yerong
Creek, seven trees were correlated with rainfall
three years previously: one with winter rainfall,
Proc. Linn.. Soe. N.SOW.13.1, 2010
W.S. SEMPLE AND T.B. KOEN
0-20 20-40
100
40-60
| Molong (12 trees, 7 years, 2000-2006) |
60-80 80-100
| Young (11 trees, 10 years, 1998-2007) |
Yerong Creek (18 trees, 11 years, 1997-2007) |
Proportion (“%) of trees with abundant buds
0-20 20-40
40-60
60-80 80-100
Proportion (%) of years
Figure 6. Proportion (“) of trees in each E. albens stand that produced abundant floral buds (rating > 3)
grouped by the proportion of years of observation (years with incomplete data excluded). For example,
12 trees (67% of 18 trees) at Springhurst were observed to produce abundant buds on just two or fewer
occasions (18% of 1i years). Except for Molong, bud assessments for 2007 are included.
three with winter and spring rainfalls and three with
spring rainfall. (Budding at a few other trees was also
correlated with warm-season rainfall but it was most
apparent at Young where six of the 11 trees were
positively correlated with summer rainfall one year
previously.) Only a few trees were correlated with
Proc. Linn. Soe: N.S:W., 131), 2010
rainfalls two and four years previously and for most
it was negative. Contrary to all the other sites, nine of
the 18 trees at Springhurst were positively correlated
with autumn rainfall four years previously. Budding
in most (but not all) trees therefore seemed to be
dependent on cool-season (either winter or autumn)
rainfall three or four years previously.
103
REPRODUCTIVE PHENOLOGY OF WHITE BOX
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
| Yerong Creek | | Yerong Creek |
Mean fruit abundance rating / tree
0
5 SS
4 | Springhurst
3
2
"4997 1998 1999 2000 2001 2002 2003 1998 1999 2000 2001 2002 2003 "004 2008 2006 2005 2006
Figure 7. Mean abundance ratings (0-5) for capsules over time at four stands of E. albens. No distinc-
tion is made between immature (usually the main component on peaks and steeply rising limbs on the
graphs) and over-mature capsules (usually the main component towards the ends of falling limbs on
each graph).
DISCUSSION early budding at a few trees at Molong (and again
in March 2007 and 2008; Semple and Koen 2008).
Budding and flowering times Buds were at a maximum by early summer and
Floral (pin) buds were usually firstevidentaround bud abundance ratings rarely declined prior to the
November — apart from some unusual occurrences of | commencement of flowering. Even so, bud shedding
104 Proc. Lmn.. Soc. N-SOW. AS 2010
W.S. SEMPLE AND T.B. KOEN
= Co
£ |
eae ie | (c) Yerong Creek |
o | : C |
@ 500 | ye | |
400 + | el | =
300 4 - i is |
rie i] Ss os es oa f
100 4 7 | | | | "
| oF | par a aso pecs : cates 22 ea Coe ie i i | 700
| _ m (9) Springhurst | | eas
! ie - | | | - 500
7 fe | ee
| I + 300
! | oo 1) El mf 200
| | + 100
is & ‘¢
T T T i ‘
SP SS gr P
SP er Fs
Figure 8. Cool (March to August) and warm (September to February of the following year) season rain-
fall from stations near the four E. albens monitoring sites. Seasonal data derived from monthly inter-
polations (as per Jeffrey et al. 2001) and long term means (thickened lines) from incomplete Bureau
of Meteorology data: Molong (1884-2006), Young (1871-1991), Yerong Creek (1885-2007), Springhurst
(1900-2007).
was probably common as has been reported for
eucalypts (Florence 1996) and for E. albens at Cowra
(Semple et al. 2007) but was not usually detected by
the relatively coarse abundance rating scale used in
this study. Flowering generally occurred from March
to November in the year following budding.
The first occurrence of buds and the flowering
period were consistent with previous observations
by Clemson (1985) and Semple et al. (2007) but the
flowering period was inconsistent with observations
by others, e.g. mid/late summer to winter, or autumn
to winter (see Table 2). Summer flowering is possible
as was demonstrated by a few trees at Molong (though
few flowers were produced and flowering did not
Proc. Linn. Soc. N.S.W., 131, 2010
extend beyond January) and for two trees at Young in
2003 (when their main flowering period commenced
in February). As some of the reports of an earlier
flowering period, i.e. between summer and winter,
predate the early 1990s, is it possible that the
flowering period has changed since c.1990 — perhaps
in response to increased frequencies of years of
below-average rainfall (e.g. Fig. 8) or even higher
temperatures in recent times. Without access to the
original observations, it is difficult to establish but the
possibility of a later and longer flowering period in
recent times cannot be ruled out.
Leigh’s (1972) report of a longer flowering
period in NSW compared to southern Queensland
105
REPRODUCTIVE PHENOLOGY OF WHITE BOX
(a) Molong (b) Molong
[winter -3] (n=7) r= 0.81 P=0.03 [winter -4] (n=7) r= -0.78 P=0.04 2
|
® ©
2 44 < 4
ie)
E e
a3 3
o
= |
Fe |
° 2 | e 2
© |
3 1 e 1
0) - : 0)
0) 0) 100 200 300 400
(c) Young (d) Young
[winter -3] (n=10) r= 0.69 P=0.03 _[winter -4] (n=10) r= -0.64 P=0.05,
2 e e
5S 4 a 4
=
B 34
oO
no)
—
ea} a) |
&
£
Soh ti
<x
o+ .
fs) 100 200 300 400 i)
(e) Yerong Creek (f) Yerong Creek
[winter -3] (n=11) r= 0.55 P=0.08 [winter -4] (n=11) r= -0.54 P=0.09,,
T
3
5 - 4
z °
3B 3
o
= e
a 2
- Tig Nee
oO
@ = 1
x e @
. = - LQ
i) 100 200 300 400
(g) Springhurst (h) Springhurst
[winter -3] (n=11) r= 0.60 P=0.05 [winter -4] (n=11) r= 0.20 P=0.95 p
oO
e 44 4
S ® 6
Ss
= 3 e e 2
no}
=
ri 2
» e
© e
ie 1
a e @
®
of e-<« eo — 0
0) 400 0) 100 200 300 400
Rainfall (mm) Rainfall (mm)
Figure 9. Correlations between maximum mean annual bud abundance and winter (June-August) rain-
fall 3 and 4 years previously at four E. albens monitoring sites. Number of years of data indicated by
106 Proc. Linn. Soc. N.S.W., 131, 2010
W.S. SEMPLE AND T.B. KOEN
suggested that it may be longer in the south, e.g.
at Springhurst, but this was not evident in the data,
albeit limited by only two peak flowering periods in
that stand. Nor was it evident in Stelling’s (1998a, b)
report for southern NSW (Table 2).
Temporal and spatial variation in flowering
Variable flowering periods and _ intensities
between individual eucalypts in a stand in any one
year is well known and has been attributed variously
to tree age/size, health and probably genotype as well
as local variations in elevation, soil types and moisture
availability (House 1997, Wilson and Bennett 1999).
As indicated by the ranges of DBHs (Table 1), trees
of variable size and presumably age were present
in each stand but the association between DBH and
the frequency of abundant budding was generally
weak. Elevation, soil type and moisture availability
appeared to be relatively uniform in each stand,
except for the hilltop stand at Young where elevation
_ varied by ~2 m. As budding intensity varied (a)
between trees in each stand in the one year and (b)
between individuals across years, e.g. some budded
abundantly in consecutive years whereas others did
not, prior rainfall alone cannot explain flowering in a
stand. If it did, then all trees would flower (or produce
buds) in a similar manner each year.
Nevertheless prior rainfall is important for tree
health and its varying occurrence and abundance
would be expected to have varying effects on
the production of new leaves and reproductive
structures. For example, Porter (1978) in attempting
to explain correlations between previous rainfall (and
temperature) and honey production (= flowering
intensity in a stand) from E. tricarpa (with similar
phenology to E. albens), noted that leaf growth was
favoured by wet summers but not by cool wet winters
- though stored water from the latter favoured growth
of floral buds in the following spring.
The data presented here indicate that individual
(and sometimes groups of) trees responded differently
to the same rainfall cues — except perhaps in those
years when most trees budded abundantly (e.g.
Fig. 4). This was supported by the examination of
correlations between bud abundance and previous
rainfall: bud abundance in some trees was not
correlated with prior rainfall (at least in the previous
five years) whereas other trees in the same stand
were correlated with differing rainfall events. Even
so, there were some broad correlations between
mean bud abundance in a stand and previous rainfall
e.g. between winter rainfall three years previously
(positive) and four years previously (negative) in
the two northern-most stands but these associations
Table 2. Flowering periods of Eucalyptus albens as reported by various authors.
Flowering period Area
Late summer and sometimes SE Australia
into winter
January to June SE Australia
February to July central western NSW
SE Australia
March to May
April to July (Qld) or August SE Australia
(NSW)
May to September southern NSW
Autumn to late spring near Cowra, NSW
April to November SE Australia
Proc. Linn. Soc. N.S.W., 131, 2010
Source
Kelly et al. 1977
Costermans 1983
Schrader 1987
Brooker and Kleinig 1990; Boland ef
al. 1992; Nicolle et al. 1994
Leigh 1972
Stelling 1998a, b
Semple et al. 2007
Clemson 1985
107
REPRODUCTIVE PHENOLOGY OF WHITE BOX
did not extend to stands further south (Fig. 9) where
mean bud abundance was correlated with other
previous rainfall occurrences. Varying genotypes
within and between stands would seem to be the
mostly likely explanation for these results; though
phenotypic variation due to (undetected) fine-scale
variation in resource availability cannot be ruled out.
Nevertheless, the presence of such variation would
increase the likelihood of floral bud and hence, seed
production in at least a few trees in each stand in most
years.
The role of flowering (and seeding) in seedling
recruitment of woodland eucalypts
The availability of a seedbank is only one of a
number of factors that affect seedling recruitment. The
success of seedbed-manipulation experiments over a
number of years in the eucalypt woodland belt (e.g.
Semple and Koen 1997, Lawrence et al. 1998, Geeves
et al. 2008) suggests that sufficient and timely rainfall
for germination and seedling establishment is not a
rare occurrence. However, unlike parts of Victoria,
seedling recruitment of woodland eucalypts is rarely
observed in NSW. For the most part, this is probably
due to the absence of a seedbed that provides exposed
mineral soil and reduced herbaceous competition — a
consequence of relatively high fertility soils (Beadle
1981) and groundstoreys that are often dominated by
exotic species (Prober 1996) in the box (e.g. E.albens
and E. melliodora) woodlands of central and southern
NSW. Though appropriate seedbeds can be deliberately
(or accidentally) prepared, e.g. by applying herbicides
or cultivating near trees, their ‘natural’ occurrence
is largely dependent on high intensity grazing (e.g.
Curtis and Wright 1993), drought (e.g. Curtis 1990)
or fire (e.g. Cluff and Semple 1994, Semple and
Koen 2001) though in the latter case, exotic species
if present, rapidly recolonise negating any initial
benefits for the eucalypt seedling. Nevertheless, when
rainfall, seedbed and other favourable conditions do
coincide, the on-going availability of seed, even if in
small amounts in a few trees, is critical for successful
recruitment. A case in point is the Molong site that
was burnt in late 2006. Though the developing 2006
seed crop was destroyed, a small amount of seed
was present from earlier (2004?) flowerings (Fig. 7)
and this yielded some seedlings beneath a few trees
(Semple and Koen 2008). Despite suboptimal rainfall,
most of these seedlings were still alive in early 2009 —
probably due to the localised absence of competition
from exotic herbage.
Predicting the future?
The view expressed by the letter-writer at the
108
start of this paper implies that flower abundance is an
indicator of some future meteorological event. Such
views are not uncommon, e.g. as reported by Duff
(2007) for observations of box trees near Jeparit in
Victoria. Results presented above suggest that bud (or
flower) abundance did not provide much information
on past, leave alone future rainfall events.
CONCLUSIONS
In general, floral (pin) buds appeared in November
and flowers were produced during the following
March to November. Flowers were produced by at
least a few trees in each stand each year except for
the southern-most stand. However, the frequency of
abundant budding, when most or all of the adult trees
flowered abundantly, declined from about 4.3 years
in 10 in the northern-most stand to two years in 10
in the south. For each tree stand, these occurrences
were important for maintaining its aerial seedbank.
Without replenishment, capsule abundance was low
after two to three years.
However, the production of reproductive
structures in individual trees was often at variance to
the stand ‘average’. In terms of the first appearance of
floral (pin) buds, it could be as early as March (rather
than the November ‘average’). Flowering in some
trees commenced as early as February (compared to
the March ‘average’) or did not finish until January
(compared to the November ‘average’). Variations
such as these were usually evident in a few trees,
particularly those at Molong, suggesting a degree of
‘plasticity’ in populations at the centre of the north-
south distribution of E. albens.
Unlike average bud abundance in tree stands,
where a high abundance year was always followed
by a year of low abundance, some individual trees
budded abundantly each year over periods ranging
from two to four years. Individual differences such
as these suggest — contrary to our suggestion from an
earlier but shorter (1995-1999) observation period at
Cowra (Semple et al. 2007) — that prior rainfall in a
particular season is not a general determinant of bud
(flower) abundance, except perhaps in those years
when all trees flower abundantly. Such variability may
have positive benefits for successful reproduction in a
variable climate such as Australia’s.
ACKNOWLEDGEMENTS
Thanks to Jeff Bradley for preparing Figure 1, Sue ~
Briggs for her constructive comments on an early version
Proc. Linn. Soc. N.S.W., 131, 2010
W.S. SEMPLE AND T.B. KOEN
of the MS, anonymous referees for their comments on a
more recent version; and Justin Hughes for facilitating the
collation of rainfall data. The project was funded by the
NSW Department of Land and Water Conservation and its
various reincarnations.
REFERENCES
Allcock, K.G. (2002) Effects of phosphorus on growth
and competitive interactions of native and introduced
species found in White Box woodlands. Austral
Ecology 27, 638-646.
Beadle, N.C.W. (1981). The Vegetation of Australia.
Cambridge University Press: Melbourne.
Boland, D.J., Brooker, M.I.H., Turnbull, J.W. and Kleinig,
D.A. (1980) Eucalyptus Seed. CSIRO: Australia.
Boland, D.J., Brooker, M.I.H., Chippendale, G.M.., Hall,
N., Hyland, B.P.M., Johnson, R.D., Kleinig, D.A.
and Turner, J.D. (1992) Forest Trees of Australia (4th
edition). Thomas Nelson / CSIRO: Melbourne.
Brooker, M.I.H. and Kleinig, D.A. (1990) A Field Guide
to the Eucalypts. Volume 1. South-eastern Australia
(revised edition). Inkata Press: Melbourne.
Burrows, G.E. (1995) Seed production in white box
(Eucalyptus albens) in the South West Slopes region
of New South Wales. Australian Forestry Journal
58, 107-109.
Clemson, A. (1985). Honey and Pollen Flora. Inkata
Press: Melbourne / Department of Agriculture New
South Wales: Sydney.
Cluff, D. and Semple, W.S. (1994) Natural regeneration:
in “Mother Nature’s’ own time. Australian Journal of
Soil and Water Conservation 7(4), 28-33.
Costermans, L. (1983) Native Trees and Shrubs of South-
eastern Australia (revised edition). Weldon: Sydney.
Curtis, D. (1990) Natural regeneration of eucalypts in the
New England region. In: Sowing the Seeds. Greening
Australia: Deakin.
Curtis, D. and Wright, T. (1993) Natural regeneration
and grazing management: a case study. Australian
Journal of Soil and Water Conservation 6(4), 30-34.
Duff, X. (2007) Old wives tales. In: Our Weather: Wet (ed
X. Duff), p. 19. Supplement to The Weekly Times
[Melbourne], 10 October 2007.
Florence, R.G. (1996) Ecology and Silviculture of
Eucalypt Forests. CSIRO: Collingwood.
Geeves, G., Semple, B., Johnston, D., Johnston, A.,
Hughes, J., Koen, T. and Young, J. (2008) Improving
the reliability of direct seeding for regeneration in
the Central West of New South Wales. Ecological
Management & Restoration 9, 68-71.
Hill, K.D. (1991) Eucalyptus. In: Flora of New South
Wales Volume 2 (ed G.J. Harden), pp. 76-142. NSW
University Press: Kensington.
House, S.M. (1997) Reproductive biology of eucalypts.
In: Eucalypt Ecology: Individuals to Ecosystems
(eds J.E. Williams and J.C.Z. Woinarski), pp. 30-55.
Cambridge University Press: Melbourne.
Proc. Linn. Soc. N.S.W., 131, 2010
Jeffrey, S.J., Carter, J.O., Moodie, K.B. and Beswick,
A.R. (2001) Using spatial interpolation to construct
a comprehensive archive of Australian climatic data.
Environmental Modelling and Software 16, 309-330.
Kelly, S., Chippendale, G.M. and Johnston, R.D. (1977)
Eucalypts Vol. 1. Nelson: Melbourne.
Lawrence, J., Semple, W.S. and Koen, T.B. (1990)
Experimental attempts at encouraging eucalypt
regeneration in non-native pastures of northern
Victoria and central western NSW. Proceedings of
the Linnean Society of NSW 119, 137-154.
Leigh, J.H. (1972) Honey and beeswax production in
semi-arid and arid Australia. In: The Use of Trees
and Shrubs in the Dry Country of Australia (ed
N. Hall), pp. 264-283. Australian Government
Publishing Service: Canberra.
Nicolle, D., Nicolle, B. and French, M.E. (1994) Eucalypts
of the Mudgee District. F & N Eucalypt Publications:
Morphett Vale.
Porter, J.W. (1978) Relationships between flowering
and honey production of red ironbark, Eucalyptus
sideroxylon (A. Cunn.) Benth., and climate in the
Bendigo district of Victoria. Australian Journal of
Agricultural Research 29, 815-829.
Prober, S.M. (1996) Conservation of the grassy white
box woodlands: rangewide floristic variation and
implications for reserve design. Australian Journal
of Botany 44, 57-77.
Prober, S.M, Thiele, K.R. and Lunt, I.D. (2002)
Identifying ecological barriers to restoration in
temperate grassy woodlands: soil changes associated
with different degradation states. Australian Journal
of Botany 50, 699-712.
Schrader, N. (ed) (1987) The Flora and Fauna of the
Parkes Shire. Parkes Naturalist Group: Parkes.
Semple, W.S. and Koen, T.B. (2001) Growth rate and
effect of sheep browsing on young eucalypts in an
anthropogenic Themeda grassland. The Rangeland
Journal 23, 182-193.
Semple, W.S. and Koen, T.B. (1997) Effect of seedbed
on emergence and establishment from surface
seeded and direct drilled seed of Eucalyptus spp.
and Dodonaea viscosa. The Rangeland Journal 19,
80-94.
Semple, W.S. and Koen, T.B. (2003) Effect of pasture type
on regeneration of eucalypts in the woodland zone of
south-eastern Australia. Cunninghamia 8, 76-84.
Semple, B. and Koen, T. (2008) A good time for a fire? A
note on some effects of wildfire on a Grassy White
Box Woodland. The Victorian Naturalist 125, 160-
165.
Semple, W.S., Koen, T.B. and Henderson, J. (2007)
Seed fall and flowering in white box (Eucalyptus
albens Benth.) trees near Cowra, New South Wales.
Australian Forestry 70, 242-252.
Stelling, F. (ed.) (1998a) South West Slopes Revegetation
Guide. Murray Catchment Management Committee
/ Department of Land and Water Conservation:
Albury.
109
REPRODUCTIVE PHENOLOGY OF WHITE BOX
Stelling, F. (ed.) (1998b) Revegetation Guide for the
Riverina Highlands. Murray Catchment Management
Committee / Department of Land and Water
Conservation: Albury.
Wilson, J. and Bennett, A.F. (1999) Patchiness of a floral
resource: flowering of Red Ironbark Eucalyptus
tricarpa in a Box and Ironbark forest. The
Victorian Naturalist 116 (2), 48-53.
110 Proc. Linn. Soc. N.S.W., 131, 2010
The Early Devonian Trilobite Craspedarges from the Winduck
Group, Western New South Wales.
LAWRENCE SHERWIN! AND N. Stmone MEAKIN?
'Geological Survey of New South Wales, Locked Bag 21, Orange, New South Wales 2800 (lawrence.
sherwin@industry.nsw.gov.au); “Geological Survey of New South Wales, PO Box 344, Hunter Region Mail
Centre, New South Wales 2310 (simone.meakin@industry.nsw.gov.au)
Sherwin, L. and Meakin, N.S. (2010). The Early Devonian trilobite Craspedarges from the Winduck
Group, western New South Wales. Proceedings of the Linnean Society of New South Wales 131, 111-
118.
Specimens of the lichid trilobite Craspedarges wilcanniae Giirich from the Early Devonian Winduck
Group in “The Meadows’ area, near Cobar, in western New South Wales, enable a revised description and
a neotype to be designated to replace types destroyed during World War II.
Manuscript received 18 January 2010, accepted for publication 26 May 2010.
KEYWORDS: Cobar, Craspedarges, Early Devonian, Lichidae, trilobites, western New South Wales,
Winduck Group.
INTRODUCTION
In ‘The Meadows’ area (Figure 1), south—west
of Cobar in western New South Wales, the Early
Devonian (Lochkovian) lichid trilobite Craspedarges
wilcanniae occurs in the Winduck Group (Glen
1987), a unit within the widely distributed Cobar
Supergroup. The stratigraphy and brachiopod faunas
of this area have been described elsewhere (Sherwin
1992, 1995) and on a broader scale the structural
setting has been described by Glen (1990). Geological
mapping in this particular area was handicapped
by poor outcrop but the favoured interpretation is
that the Winduck and Amphitheatre Groups have
an interfingering relationship (Figure 2), with the
Winduck Group sedimentation continuing for a longer
period. Trilobites have not been reported previously
from this area, the nearest occurrences in the Cobar
Supergroup being in the vicinity of Cobar (Baker et
al. 1975, Fletcher 1975), 60 kilometres north-east
of “The Meadows”. Ebach and Edgecombe (1999)
described a new species of the proetid Cordania
from the vicinity of “The Bluff’, south of Cobar,
in the Biddabirra Formation (Amphitheatre Group)
which underlies the Winduck Group. Fletcher (1975)
also described several other species of trilobites
from the vicinity of Cobar and several localities
north-east of Nymagee where Webby (1972) had
noted an Encrinurus occurrence. From that same
area, Landrum and Sherwin (1976) described a new
proetid, Warburgella (Anambon) jelli, regarded by
Yolkin (1983) as a junior synonym of the Eurasian
species Warburgella tcherkesovae Maximova and
Warburgella waigatschensis (Tschernyschev and
Yakovlev, 1898). Strusz (1980) reviewed the species
of Encrinurus described by Fletcher and regarded
the specific attributions as doubtful because of the
poor preservation. The stratigraphy of the Nymagee
localities has been described by Felton (1981). The
lichid trilobite Craspedarges wilcanniae Giirich,
found at several localities within the Winduck Group,
was described from erratics, believed derived from
the Cobar Supergroup, in Cretaceous sediments at
White Cliffs (Giirich 1901) about 230 kilometres
north-west of “The Meadows” (Figure 1).
Several genera of trilobites are represented in
“The Meadows” district but only the lichid species
is described here. The encrinurids occur in pinkish
mudstones of the Late Silurian to Early Devonian
Amphitheatre Group and are generally complete,
although fine details are not well preserved. In
the Winduck Group probable Gravicalymene 1s
associated with Craspedarges but is otherwise too
poorly preserved to warrant description and proetids
are represented by a nondescript pygidium.
AGE OF THE FAUNA
The brachiopods associated with Craspedarges
wilcanniae indicate an Early Devonian (Lochkovian)
age (Sherwin 1995). The only other recorded species
of Craspedarges, C. superbus, was described from
CRASPEDARGES (TRILOBITE) FROM WESTERN NSW
y
Lo aw The Meadows
SU =
roy
a on
| Buckwat?—
REFERENCE
Cainozoic —— Fault
== Strike and dip
Mid-Late Devonian Xe Fold axis
[::"], Mulga Downs Gp mh. abate
Early Devonian ———~ __ Drainage
|__|} Winduck Gp
[| Amphitheatre Gp
eae
falipedai if \
| {
/
White Cliffs {
Cobar /
7 Broken Hill | -Nymagee /
| ewcastle
any SYDNEY
Upn #Wollongong
CANBERRA f
\ re
Ni aes G “he
ee et ge SO |
ken Ie
ret
4 re ee AE \
| Pipa NSW
| bs a vic. /
U/ Tas
Figure 1. Locality diagram showing places mentioned in text, fossil localities and geological sketch
map, modified from Rose (1965).
the ‘Gedinnian to Emsian or early Eifelian’ Fukuji
Series in Japan by Kobayashi and Hamada (1977a,
b), although the generic identification was queried
by Thomas and Holloway (1988). Lichid trilobites
have been described from Early Devonian (Pragian—
Emsian) limestones in New South Wales (Edgell
1955; Chatterton 1971; Chatterton et al. 1979;
Edgecombe and Wright 2004) and quartzose clastics
in Victoria (Gill 1939; Holloway and Neil 1982) but
all belong to the genus Acanthopyge except for one
doubtful reference to Terranovia from New South
Wales (Chatterton and Wright 1986).
I
SYSTEMATIC PALAEONTOLOGY
Morphological terms, unless otherwise specified,
are as defined in the Treatise on Invertebrate
Paleontology (Moore, ed. 1959), supplemented with
lichid morphology of Thomas and Holloway (1988)
except that we do not regard the occipital ring as
part of the glabella.. All specimens are stored in the
collections of the Geological Survey of New South
Wales at Londonderry in western Sydney. External
moulds were studied using latex casts and all
specimens, whether casts or originals, were whitened
with MgO for photography. Actual specimens were
blackened with water colour before application of '
MgO.
Proc. Linn. Soc. N.S.W., 131, 2010
L. SHERWIN AND N.S. MEAKIN
Middle—Late
Emsian
=
=
5 =| Pragian
Sia
iy | Ww [or eee
oO SSE eernes : a
Winduck me a
Group BUTE
Lochkovian ee
Te =e = Amphitheatre :
SS =a Group
~~ NBT M12
|
SILURIAN Winduck
Shelf Cobar Basin
Figure 2. Stratigraphic relationships in “The Meadows” dis-
trict, modified from Glen (1987), showing approximate strati-
graphic position of trilobite localities. Craspedarges wilcanni-
ae occurs at localities TM56b and TM65. Encrinurus occurs
at localities NB1 and TM312. In this area it has not been pos-
sible to recognise formations within the Amphitheatre and
Winduck Groups.
Family LICHIDAE Hawle and Corda, 1847
Subfamily TROCHURINAE Phleger, 1936
Craspedarges Giirich, 1901
third subparallel to rachis; abaxial ends
of pleurae continued beyond border as
tapered spines with circular cross sections;
rachis parallel sided for approximately
one third length of pygidium, remainder
tapered and continued beyond border as
terminal spine flanked by a pair of border
spines.
Remarks
The types of this genus are believed
to have been destroyed with the remainder
of Giirich’s collection, housed originally
in Breslau (now Wroclaw), when
Hamburg was bombed during World
War II. Although a significant part of the
collection survived the war, there is no
trace of the types of Craspedarges or even
the associated brachiopods (J. Dzik, pers.
comm.).The search described by Thomas
and Holloway (1988) was repeated as well
as extended to the Geological Survey of
New South Wales collections without any
success. This redescription is based upon
material found in situ in sandstones of
the Winduck Group. Giirich’s types came
from erratic boulders, as noted above,
but the exact source, or sources, of the
etratics is unknown, there being very little
pre—Quaternary outcrop between White
Cliffs and ‘The Meadows’, although the erratics are
comparable in lithology and faunal content (Dun
1898) with the Winduck Group.
Because of doubts about the source of the
Type species
Craspedarges wilcanniae Girich, 1901
erratics it is necessary to establish that the lichids
from the Winduck Group are truly Craspedarges.
Giirich’s material consisted of an internal mould of an
Diagnosis (revised)
Trochurine with very globose cranidium; anterior
border wide and gently convex in section (sag.),
becoming flatter near suture; longitudinal furrows
shallow posteriorly, much deeper anteriorly including
in front of S1 and subparallel for most of length from
posterior edge of cranidium, diverging anteriorly to
join border furrow; S1 deep behind bullar lobes, weak
between longitudinal furrows; portion of L1 between
longitudinal furrows much lower than occipital ring
and median lobe but approximately the same width
(trans.) as the occipital ring. Pygidium approximately
as wide as long with narrow well developed raised
border; rachis approximately one third the maximum
width of the pygidium; first pair of pleurae backwardly
flexed, second less so but more inclined to rachis,
incomplete cranidium and three fragmentary moulds
of ventral surfaces of the pygidium. The cranidium,
except for some flattening indicated by a line
drawing of the profile, matches the Winduck Group
material. Matching the pygidia is difficult because
the one pygiditum known from the Winduck Group
has more or less uniformly slender marginal spines
preserved whereas two (Giirich, pl. 18, figures 6 and
8) of Giirich’s specimens have comparatively short
and wide spines. These two particular specimens are
very fragmentary and it is not at all certain that they
belong to the same species, ie., C. wilcanniae. The
remaining fragment illustrated by Giirich (pl. 18,
figure 7) is of the posterior margin and is reconcilable
to a greater extent with the Winduck Group specimen.
Giirich’s specimens are illustrated by drawings only
so that there is a possibility that the figures are not
Proc. Linn. Soc. N.S.W., 131, 2010
3)
CRASPEDARGES (TRILOBITE) FROM WESTERN NSW
necessarily an accurate representation of the original
specimens, especially his diagrammatic sketch of a
flattened and incomplete cranidium (pl. 20, figure
20). The illustration in the trilobite Treatise (Moore
1959, figure 396—6a) is a line drawing that does not
correspond with either of Giirich’s sketches but seems
to be based upon a composite of the two. The cephalic
profile in the Treatise (figure 396—6b) is clearly
copied from Giirich (figure 1a) but the anterior border
has been changed from planar to slightly concave and
the figure generally flattened. In this paper (figure
3, A and B) a slightly flattened cranidium has been
placed alongside the comparatively undeformed
neotype to show the distorted anterior border
resembles the Treatise illustration. The shading in
Giirich’s illustration (pl. 18, figure 1) suggests that
some convexity remains in the left side of the anterior
border.
Craspedarges is closely related to Richterarges,
as noted by Thomas and Holloway (1988), the major
differences being the more prominent anterior border
and much deeper anterior part of the longitudinal
furrows. A slight midlength expansion in the
median lobe of Richterarges has no analogue in the
corresponding part of Craspedarges where the sides
of the median lobe are straight. The pygidium of
Richterarges has only two distinct pleurae compared
with three in Craspedarges. Thomas and Holloway
also postulated that Craspedarges was derived
from Richterarges in about Late Silurian to Early
Devonian time, which accords with the age of the
Winduck Group. However. the pygidial segmentation
in Craspedarges is less effaced than Richterarges,
suggesting that it departed earlier from the ancestral
hemiargid stock.
Pollit et al. (2005) carried out a cladistic study
and Bayesian analysis of the Family Lichidae but
excluded Craspedarges from consideration because
of its poorly known morphology; they did recognise
that it is closely related to the group represented
by Acanthopyge, Akantharges, Ceratarges and
Borealarges and in other respects to the group
containing Richterarges and Terranovia.
Craspedarges wilcanniae Giirich, 1901 (Figure 3)
1901 Craspedarges wilcanniae Girich, p. 532-538,
pl. 18, figures 1, 6—8; pl. 20, figure 20.
Neotype
MME 31377(5) a cranidium lacking the
postero—lateral extremities.
114
Neotype locality
TM 56b, Winduck Group, Early Devonian
(Lochkovian).
Other material
MMEF 31333 anterior of cranid‘um: MMF 31334
posterior half of cranidium; MMF 31399 and 31400
poorly preserved cranidia; MMF 31377(10) and (11)
hypostomes; MMF 31398 incomplete pygidium. The
numbers in brackets refer to individual specimens on
slabs with numerous fossils.
Other localities
TM 65, Winduck Group (MMF 31399 only).
Diagnosis
Craspedarges with 1L undivided between
longitudinal furrows.
Description
The cranidium is very strongly convex, almost
globose. The border is very distinct and anteriorly
convex in section (sag.), being broadest near the
anterior and posterior ends of the suture. The border
furrow is narrow, except at the genal angles, and
well defined. The rachial furrows are indistinct on
the posterior border and effaced on the postero—
lateral cranidial lobe between the palpebral lobe
and posterior border furrow. The occipital ring is
poorly defined laterally because of the weak posterior
rachial furrows, but is clearly differentiated from 1L
by the occipital furrow. The longitudinal furrows are
weak between the posterior margin and S1 but deep
anteriorly and sub—parallel along the inner sides of the
bullar lobes. The median part of 1L is well marked by
the longitudinal furrows and comparative depression
among otherwise inflated lobes but the lateral ends
are lost in the undifferentiated postero—lateral
cranidial lobes. The bullar lobes are clearly defined
by the circumscribing furrows. The median lobe is
the most inflated part of the cranidium and very wide
anteriorly, though the antero—lateral extremities do not
overlap the bullar lobes. The surface is covered with
small pointed tubercles that are finer on the border.
[The perforations on some tubercles are believed
to be bubbles in the latex cast and are irregular in
distribution.] The free cheeks are unknown.
The hypostome is wider than long although
the posterior border is incomplete on both specimens.
The posterior lobe is narrow (sag.) and crescentic in
shape compared with the larger subquadrate anterior
lobe. The surface of at least the median body is
Proc. Linn. Soc. N.S.W., 131, 2010
L. SHERWIN AND N.S. MEAKIN
Figure 3. Craspedarges wilcanniae Giirich; A, A’ MMF 31377(5) neotype, stereo pair of latex cast
of exterior of incomplete cranidium; B MMF 31399 latex cast of exterior of flattened incomplete
cranidium showing impact on anterior border; C, C?’ MMF 31334 stereo pair of latex cast of
exterior of posterior part of cranidium; D MMF 31377(11) latex cast of interior of hypostome;
E-F MMF 31377(10) latex casts of interior and exterior of hypostome, E interior, F, F’ stereo
pair of incomplete exterior; G MMF 31398 latex cast of incomplete pygidium.
ornamented with tubercles finer but otherwise
comparable with those on the cranidium.
No thoracic segments of this species are
known.
The only pygidium is incomplete at its
anterior edge and the rings are not preserved on the
prominent rachis. The posterior edges of the three
pleurae form well defined ribs in the pleural fields, the
ribs on the second and third pleurae being continued
beyond the well defined raised border as robust
Proc. Linn. Soc. N.S.W., 131, 2010
spines. The very poorly preserved internal mould,
counterpart to the exterior in Figure 3G, shows that
the first pleura is also continued beyond the border
as a marginal spine of uncertain length. The internal
mould also shows a short, comparatively broader
Spine corresponding to the anterior edge of the second
pleura, making a total of five pairs of marginal spines.
The pair flanking the terminal spine are in the position
that would correspond to a fourth pair of pleurae. The
surface is covered with irregularly distributed and
115
CRASPEDARGES (TRILOBITE) FROM WESTERN NSW
widely spaced granules. The doublure is unclear in
extent but is approximately as wide as the border.
Dimensions
Because of the fragmentary preservation
some of the dimensions have been extrapolated by
doubling measurable half widths.
length width
(mm) (mm)
MMF 31377(5) cranidium 9.0 9.5
MME 31334 cranidium (posterior) 12.5
MMF 31398 pygidium (ex spines) 10.5 10.0
Remarks
The reasons for assuming that these specimens
are truly conspecific with Giirich’s originals are
discussed under the generic remarks. The only
other species assigned to this genus, Craspedarges
superbus Kobayashi and Hamada (1977a) from
Japan, was questionably assigned to Richterarges
by Thomas and Holloway (1988), although this
decision was influenced by the poorly known
morphology of Craspedarges wilcanniae. The extra
pair of pleural segments and five pairs of marginal
spines on the pygidium described by Kobayashi and
Hamada (1977a) is in agreement with Craspedarges
wilcanniae, the main distinction being that S1 in
Craspedarges superbus is not discrete but instead
merges medially with the occipital furrow. The age of
Craspedarges superbus is imprecise, Kobayashi and
Hamada (1977b) giving an age range from Gedinnian
to early Eifelian. The earlier limit accords with the
age of Craspedarges wilcanniae and the Winduck
Group.
ACKNOWLEDGMENTS
This paper is part of a PhD carried out by L. Sherwin at
Macquarie University under the supervision of J.A. Talent
and R. Mawson. With respect to mapping in ‘The Meadows’
area we thank fellow Geological Survey of NSW staff John
Byrnes, John Chapman, Gary Dargan, Dick Glen, Hervey
Henley, Dave Jones, Dennis Pogson and John Watkins;
cartography is by Cheryl Hormann. David Holloway
(Museum Victoria) provided copies of relevant trilobite
publications. The following are thanked for searching for
Giirich’s type material: Dr. J. Dzik (Zaktad Paleobiologii,
Warsaw), Mrs J. Poleska (Muzeum Geologiczne, Wroclaw
University), Dr G.K.B. Alberti and Dr W. Weitschat (both
Hamburg University). Published with the permission of the
Director, Geological Survey of New South Wales, Industry
and Investment NSW.
116
REFERENCES
Baker, C.J., Schmidt, B.L. and Sherwin, L. (1975).
Revised stratigraphy of the Cobar-Gunderbooka
area. New South Wales Geological Survey—Quarterly
Notes, 20, 1-4.
Chatterton, B.D.E. (1971). Taxonomy and ontogeny of
Siluro—Devonian trilobites from near Yass, New
South Wales. Palaeontographica A 137, 1-108.
Chatterton, B.D.E., Johnson, B.D. and Campbell, K.S.W.
(1979). Silicified Lower Devonian trilobites from
New South Wales. Palaeontology 22, 799-837.
Chatterton, B.D.E. and Wright, A.J. (1986). Silicified
Early Devonian trilobites from Mudgee, New South
Wales. Alcheringa 10, 279-296.
Dun, W.S. (1898). Notes on the fauna of the Devonian
boulders occurring at the White Cliffs opal fields.
Geological Survey of New South Wales—Records, 5,
160-174.
Ebach, M.C. and Edgecombe, G.D. (1999). The Devonian
trilobite Cordania from Australia. Journal of
Paleontology, 73, 431-436.
Edgecombe, G.D. and Wright, A.J. (2004). Silicified Early
Devonian trilobites from Brogans Creek, New South
Wales. Proceedings of the Linnean Society of New
South Wales 124, 177-188.
Edgell, H.S. (1955). A Middle Devonian lichid trilobite
from south-eastern Australia. Palaeontologische
Zeitschrift, 29, 136-145.
Felton, E.A. (1981). Geology of the Canbelego 1:100 000
sheet 8134. New South Wales Geological Survey,
Sydney.
Fletcher, H.O. (1975). Silurian and Lower Devonian
fossils from the Cobar area of New South Wales.
Records of the Australian Museum 30, 63-85, 5 figs.
Gill, E.D. (1939). The Silurian trilobite Lichas australis.
Memoirs of the National Museum of Victoria 11,
140-142.
Glen, R.A. (1987). Geology of the Wrightville 1:100 000
Sheet 8034. Geological Survey of New South Wales,
Sydney.
Glen, R.A. (1990). Formation and inversion of
transtensional basins in the western part of the
Lachlan Fold Belt, Australia, with emphasis on the
Cobar Basin. Journal of Structural Geology 12,
601-620.
Giirich, G. (1901). Uber eine neue Lichas — Art aus dem
Devon von Neu-Sud-Wales und tiber die Gattung
Lichas iiberhauft. Neues Jahrbuch ftir Mineralogie
und Paldontologie 14, 519-539.
Hawle, I. and Corda, A.J.C. (1847). Prodrom einer
monographie der béhmischen Trilobiten.
Abhandlungen Koeniglichen B6hmischen
Gesellschaft der Wissenschaften. J.G. Clave, Prague.
Holloway, D.J. and Neil, J.V. (1982). Trilobites from
the Mount Ida Formation (Late Silurian—Early
Devonian), Victoria. Royal Society of Victoria-—
Proceedings 94, 133-154.
Proc. Linn. Soc. N.S.W., 131, 2010
L. SHERWIN AND N.S. MEAKIN
Kobayashi, T. and Hamada, T. (1977a). Devonian trilobites
of Japan in comparison with Asian, Pacific and other
faunas. Palaeontological Society of Japan Special
Papers 20, 202 pp.
Kobayashi, T. and Hamada, T. (1977b). Outline of
Devonian trilobites in Japan. Proceedings of the
Japan Academy 53, 147-150.
Landrum, R.S. and Sherwin, L. (1976). Warburgella from
central New South Wales. Geological Survey of New
South Wales—Records 17, 135-146.
Maximova, Z.A. (1970). Siluriiskiye trilobiti ostrova
Vaigach. V kn.: Stratigraphia i fauna Siluriiskih
otlozhyenii Vaigacha, 195-209
Moore, R.C. (ed. 1959). Treatise on invertebrate
paleontology, vol. O (1) (Arthropoda). Geological
Society of America and University of Kansas Press, ©
xix + 560 pp.
Phleger, F.B. (1936). Lichadian trilobites. Journal of
Paleontology 10, 593-615.
Pollitt, J.R., Fortey, R.A. and Wills, M.A. (2005).
Systematics of the Trilobite Families Lichidae Hawle
& Corda, 1847 and Lichakephalidae Tripp, 1957: the
application of Bayesian inference to morphological
data. Journal of Systematic Palaeontology 3, 225-
241.
Rose, G. (1965). Barnato 1:250 000 Geological Sheet SH
55—13. New South Wales Geological Survey, Sydney.
Sherwin, L. (1992). Siluro—Devonian biostratigraphy of
central New South Wales. Geological Survey of New
South Wales—Quarterly Notes 86, 1-12.
Sherwin, L. (1995). Siluro—Devonian brachiopods from
the Amphitheatre and Winduck Groups (Cobar
Supergroup), western New South Wales. Memoir of
the Association of Australasian Palaeontologists 18,
61-96.
Strusz, D.L. (1980). The Encrinuridae and related trilobite
families, with a description of Silurian species from
southeastern Australia. Palaeontographica A 168,
1-68.
Thomas, A.T. and Holloway, D.J. (1988). Classification
and phylogeny of the trilobite order Lichida. Royal
Society of London—Philosophical Transactions B 321,
179-262.
Tschernyschev, F. and Yakovlev, N. (1898). Fauna
izvyestnyakov misa Grebyeni na Vaigachye 1 r.
Nyehvatovoi na Novoi Zyemlye. Izv. Geol. Komityeta
t. xvii, No. 8. [in Russian].
Webby, B.D. (1972). Devonian geological history of the
Lachlan Geosyncline. Journal of the Geological
Society of Australia 19, 99-123.
Yolkin, E.A. (1983). Regular patterns in Dechenellid
evolution and biochronology of the Silurian and
Devonian. Academy of Sciences of the USSR
— Siberian Branch Transactions, 116 pp. [in
Russian].
Proc. Linn. Soc. N.S.W., 131, 2010
1
CRASPEDARGES (TRILOBITE) FROM WESTERN NSW
APPENDIX
FOSSIL LOCALITIES
Grid references (GR) are from “The Meadows’ 1:100 000 topographic map. Other localities were sam-
pled using the Barnato 1:250 000 grid; the original grid reference, shown in brackets, has been retained. Un-
less otherwise stated the fossils are in sandstone beds protruding above the surrounding scree of finer, more
thinly bedded sediments or soil. All localities are within the Cobar Supergroup but in this region it has not
been possible to subdivide the Amphitheatre and Winduck Groups.
NB | GR 559 123 (Barnato 1:250 000 GR 34601015): unnamed off white fine grained quartzose sandstone
member, Amphitheatre Group.
TM 56b GR 459 008: fine grained micaceous quartz sandstone, Winduck Group.
TM 65 GR 4630 0095: fine grained orthoquartzite, Winduck Group.
TM 312 GR 505 130: pale reddish purple massive or thickly bedded siltstone exposed in gravel scrapes,
Amphitheatre Group.
118 Proc: Linn. /Soc.N:Sow., a3, 2010
Sexual Dimorphism in the Adult South African (Cape) Fur
Seal Arctocephalus pusillus pusillus (Pinnipedia: Otariidae):
Standard Body Length and Skull Morphology
C. L. Stewarpson’, T. Prvan?, M. A. MEverR? AND R. J. Rivcute**
'Botany and Zoology, Australian National University, Canberra, ACT 2601, Australia.
(Present Address, Fisheries and Marine Sciences Program Bureau of Rural Sciences, The Department of
Agriculture, Fisheries and Forestry, Canberra, ACT 2601, Australia).
"Department of Statistics, Macquarie University, NSW 2109, Australia.
*Marine and Coastal Management (MCM), Rogge Bay, Cape Town, South Africa.
*School of Biological Sciences, The University of Sydney, NSW 2006, Australia.
*Corresponding Author: Raymond J. Ritchie, School of Biological Sciences, The University of Sydney, NSW
2006, Australia, email rrit3 143 @usyd.edu.au.
Stewardson, C.L Prvan, T., Meyer, M.A. and Ritchie, R.J. (2010). Sexual dimorphism in the adult
South African (Cape) fur seal Arctocephalus pusillus pusillus (Pinnipedia: Otariidae): standard body
length and skull morphology. Proceedings of the Linnean Society of New South Wales 131, 119-140.
We examine differences in standard body length and skull morphology of male (n = 65) and female (n =
18) South African (Cape) fur seals, Arctocephalus pusillus pusillus, from the coast of southern Africa
with the aim to develop an objective method for determining the sex of fur seal skulls. Males were
found to be significantly larger than females in standard body length, with K-means cluster analysis
successfully identifying 2 relatively homogeneous groups. Principal component analysis (covariance
matrix) showed that the underlying data structure for male and female skull variables was different, and
that most of this variation was expressed in overall skull size rather than shape. Males were significantly
larger than females in 30 of the 31 skull variables. Breadth of brain case was significantly different for the
genders. Relative to condylobasal length, males were significantly larger than females in 13 of the 31 skull
variables used in the present study. These were gnathion to posterior end of nasals, breadth at preorbital
processes, least interorbital constriction, breadth at supraorbital processes, greatest bicanine breadth,
breadth of palate at postcanine | and 3, calvarial breadth, mastoid breadth, gnathion to anterior of
foramen infraorbital, gnathion to posterior border of preorbital process, height of skull at base of
mastoid and height of mandible at meatus. In males, these variables were associated with the acquisition
and defense of territory (e.g., large head size and mass; increased structural strength of the skull; increased
bite capacity). Two skull ratio parameters, breadth of braincase/condylobasal length and length of upper
postcanine row/condylobasal length were significantly higher in females compared to males. Based solely
on the skull data, mature males can be reliably distinguished from immature males and females using
both (a) Classification and Regression Tree (CART) and (b) Hierarchical Cluster Analysis. Both
approaches had difficulty in reliably distinguishing immature males from females. The Classification
and Regression Tree method was the more successful in correctly distinguishing immature males from
females.
Manuscript received | October 2009, accepted for publication 21 April 2010.
KEYWORDS: 4rctocephalus pusillus pusillus, identification of sex, multivariate analysis, Otartidae,
polygyny, Pinnipeds, principle component and cladistic analysis, sexual dimorphism, skull morphometrics,
South Africa fur seal, standard body length.
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
INTRODUCTION
Sexual dimorphism isa form of non-geographic
variation that can be generated in a species by
the process of sexual selection (Bartholomew,
1970; Alexander et al., 1979; Stirling, 1983).
Highly polygynous species such as fur seals, sea
lions and elephant seals, generally exhibit a high
degree of sexual dimorphism (Laws, 1953;
Ralls, 1977; Alexander et al., 1979; Stirling,
1983; Sirianni and Swindler, 1985; McLaren,
1993; Arnould and Warneke, 2002). Differences
in reproductive success among males of these species
are large, and competition for access to females
is intense. Selection pressure appears to favour
the development of traits that enhance male
fighting ability, including intimidating body
size, weaponry and skin thickness (Laws, 1953;
Bartholomew, 1970; Le Boeuf, 1974; Alexander et al.,
1979; McCann, 1981; Stirling, 1983).
Breeding Southern fur seals (Arctocephalus
spp.) are among the most territorial of animals,
are strongly sexually dimorphic in body size,
polygynous and gregarious (Peterson, 1968; Harrison
et al., 1968; Stirling, 1970; Bryden, 1972; Alexander
et al., 1979; Bonner, 1981; McKenzie et al., 2007).
In the southern hemisphere, breeding status male
fur seals (beachmasters) generally arrive at the
rookeries around November to establish ter-
ritories. Pregnant females arrive soon after. Once
females are present in the male’s territory, males
guard females until they come into oestrus post-
partum. Females give birth within one week of
coming ashore and then mate with the nearest
male during the short breeding (pupping/ mating)
season (Guinet et al., 1998). Males seldom leave
the territory until the breeding season is over (Rand,
1967; Stirling, 1970; Miller 1974; Peterson, 1968;
Harrison et al., 1968; Bonner, 1981). After mating,
the territorial system gradually breaks down and
males return to sea to replenish their physiological
reserves. Males do not care for their young.
When establishing territories, male fur seals
threaten each other with vocal and visual displays,
emphasising their size, to intimidate competitors
(Bonner, 1968; Stirling, 1970; Stirling and Warneke,
1971; Miller, 1974; Shaughnessy and Ross, 1980).
Much time is spent in making visual and vocal threats
to rival males and chasing them away, but fights may
develop, occasionally resulting in severe injury
or death (Rand, 1967; Stirling, 1970; Shaughnessy
and Ross, 1980; Trillmich, 1984; Campagna and Le
Boeuf, 1988).
120
Adult male fur seals are about 3 to 5 times
heavier and about 1/4 longer than adult females
(Stirling, 1983; David, 1989; Boness, 1991; Guinet et
al., 1998: Arnould and Warneke, 2002; Stewardson et
al., 2009). Large body size is in itself an intimidating
form of display to discourage rival males from
attempting an actual physical challenge and in the
event of a physical challenge is advantageous in
competitive interactions and enables breeding bulls
to remain resident on territories for longer periods
of time without feeding (Rand, 1967; Miller, 1975;
Payne, 1978, 1979; Stirling, 1970, 1983). Strong fore-
quarters, enlarged jaw and neck muscles, robust
canines, increased structural strength of the skull,
and long, thick neck hair (protective mane or wig),
also appear to be potentially advantageous in the
acquisition and maintenance of territory; quan-
titative information on these features, however, are
lacking (Miller, 1991).
Here we examine morphological differences
between skulls (n = 31 variables) of male (n =
65) and female (n = 18) South African (Cape) fur
seals Arctocephalus pusillus pusillus, from the coast
of southern Africa. Body length information was
also included in analyses where available. Where
possible, comparisons are made to the closely related
Australian fur seal Arctocephalus pusillus doriferus
(King, 1969; Brunner, 1998ab, 2000; Brunner et
al., 2002; Arnould and Warneke, 2002; Brunner et
al., 2004; Stewardson et al., 2008, 2009) and other
otarid species for which morphological data are
available such as the Steller sea lion (Ewmetopias
Jubatus) (Winship et al., 2001).
For many life history, conservation and
ecological studies it is important to be able to
determine the sex of skull material in museum
collections, skulls of animals found dead or
accidentally killed in fishing operations or killed in
other ways. Often only the skull is available. We
show that two types of multivariate analysis [(a)
Classification and Regression Tree (CART)
and (b) Hierarchical Cluster Analysis] can be used
to objectively distinguish mature male, immature
male and female skulls of the South African fur seal
(A. pusillus pusillus). By extension the approach
could be applied to other fur seals, particularly the
Australian fur seal (A. pusillus doriferus) and the
New Zealand fur seal (A. australis forsteri).
MATERIALS AND METHODS
Collection of specimens
South African (Cape) fur seals (Arctocephalus
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
pusillus pusillus) were collected along the Eastern
Cape coast of South Africa between Plettenberg
Bay (34° 03’S, 23° 24’E) and East London (33°
03°S 27 54’E), from August 1978 to December 1995
(Stewardson et al., 2008, 2009), and accessioned
at the Port Elizabeth Museum (PEM). Specimens
were collected dead or dying from the coastline and
some from accidental drowning in fishnets; none
were deliberately killed (cf. Guinet et al., 1998).
Routine necropsies were performed and biological
parameters recorded, based on recommendations
of the Committee on Marine Mammals (1967).
Animals were aged from incremental lines observed
in the dentine of upper canines (Stewardson et al.,
2008, 2009). The sample was supplemented with
measurements from 11 known-aged adult males
(animals tagged as pups) from Marine and Coastal
Management (MCM), Cape Town. The specimens
from the MCM collection have accession numbers
beginning with MCM (e.g. MCM 1809). The MCM
collection also housed 5 tag-aged adult females
and 3 tag-aged sub adult/juvenile females.
All animals considered adults had reached full
reproductive capacity, i.e., males > 8 y (Stewardson
et al., 1998; Stewardson et al., 2008, 2009) and
females = 3 y (J.H.M. David, pers. comm.). When
age was not known, males = 170 cm (Stewardson et
al., 2008, 2009) and females => 135 cm (Guinet et al.,
1998; J.H.M. David, pers. comm.) were considered
fully adult males and females and included in the
analysis as adults even if their dentition age was less
than 8 y for males. South African fur seals => 12 y
cannot be aged from counts of growth layer groups
(GLG) in the dentine of upper canines because of
closure of the pulp cavity. Estimated longevity for
male South African Fur seals is c. 20 y (Wickens, 1993;
Stewardson et al., 2008, 2009). There is much less
information on the longevity of female South African
fur seals (despite the large numbers of animals that are
shot in culling and hunting operations) but Wickens
(1993) based on zoo records concluded that females
could live to c. 30 y.
Australian male fur seals (4. pusillus doriferus)
also have a similar lifespan of about 20 years but female
Australian fur seals based on age tags are currently
known to live to well over 20 y (Amould and Warneke,
2002). Seal life spans in a range of seal species average
about 15 to 20 y for males and in excess of 20 y for
females (New Zealand fur seal (A. australis forsteri),
McKenzie et al., 2007; Antarctic fur seal (A. gazella),
Payne, 1978, 1979); Steller sea lion (Eumetopias
Jubatus), Winship et al., 2001).
Proc. Linn. Soc. N.S.W., 131, 2010
Museum records
The data set on the males used in the present study
has already been published in (Stewardson et al., 2008)
and further details can be found in Stewardson (2001).
The list of male specimens used in the present study is
shown in Appendix 1. There were 39 adult males, 24
immature sub adult males and two juvenile males only
2 years old. No standard body length measurements
were available on four (4) of the adult males (PEM
2004, PEM 2007, PEM 2013, PEM 2036) but it is
unlikely that any adult male skulls would be assigned
to the wrong sex because mature male skulls are much
larger than females and more heavily built. However,
there were no SBL measurements available on four
(4) of the immature males (PEM 2006, PEM 2009,
PEM 2010 and PEM 2014). This raises some doubts
about the certainty that these specimens were correctly
identified as males. Generally if the SBL had been
determined, the genitalia would have been available
for examination. The raw data set for the females
(18 adults, 4 juveniles and sub adults) is shown in
Appendix 2 and the means and standard deviations
in Appendix 3. All the female carcasses were complete
enough for reliable determination of their sex.
Skull variables
A total of 32 skull measurements were recorded
(Table 1). However, one of these variables, height of
sagittal crest, was not examined statistically because
there were few measurements for females and also
because we have found that sagittal crest measurements
seem to provide little useful information in male skulls
(Stewardson et al., 2008). Thus, statistical analysis
was conducted on 31 of the 32 variables. Skull
preparation and measurement procedures follow
Stewardson et al. (2008).
Statistical analyses
Six methods of analyses were employed. Firstly,
two sample t-tests (assuming equal variance) were
used to test the hypothesis that the mean value of a
skull variable was significantly different for males
and females against an appropriate alternative
MOU NESTS (als Me creole le ies a We ees
> vale) Since more than | skull variable was being
considered, the Bonferroni correction was used - the
experiment-wise error rate was divided by the total
number of tests performed (Cochran, 1977).
Secondly, K-means clustering, a non-
hierarchical cluster analysis was used to classify
observations into 1 of 2 groups based on some of the
skull variables. Observations on some of the skull
variables from both sexes were pooled so that initially
there is a single cluster with its centre as the
121
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
mean vector of the variables considered. These
observations were then assigned at random to two
sets. Step | entails calculating the mean vector of
the variables considered (centroid) for each set.
Step 2 entails allocating each observation to the
cluster whose centroid is closest to that observation.
These two steps are repeated until a stopping
criterion is met (there is no further change in the
assignment of the data points). Before doing this
all variables were standardised. Closest neighbour
(similarity) was measured using Euclidean distance
(Johnson and Wichern, 1992). The groupings of skull
variables we considered were dorsal, palatal, lateral
and mandibular. We also used k-means clustering to
classify observations into | of 2 groups using standard
body length.
Thirdly, plots of log, of each skull variable against
log, of standard body length (SBL) for the genders
were examined. ‘Robust’ regression (Huber M-
Regression) was used to fit straight lines (log y = log a
+ b log x) to the transformed data (Weisberg, 1985;
Myers, 1990).
Fourthly, principal component analysis (PCA)
was used. One useful application of PCA 1s identifying
the most important sources of variation in anatomical
measurements for various species (Jackson, 1991;
Jolliffe, 2002). When the covariance matrix is used and
the data has not been standardized the first principle
component (PC) usually has all positive coefficients
and according to Jolliffe (2002) this reflects the
overall ‘size’ of the individuals. The other PCs
usually contrast some measurements with others
and according to Jolliffe (2002) this can often be
interpreted as reflecting certain aspects of ‘shape’,
which are important to the species.
Skull measurements were recorded in the same
units; therefore a covariance matrix was used to
calculate PCs (however this gives greater weight to
larger, and hence possibly more variable measurements
because the variables are not all treated on an equal
footing). Genders were examined separately
because the grouped PCA was quite different,
in most cases, to either the separate male PCA or
female PCA.
PCA and two sample t-tests were calculated in
Minitab (Minitab Inc., Slate College, 1999, 12.23).
K-means cluster analyses for skull variables and
SBL were calculated in Minitab (Minitab Inc.,
Slate College, 1999, 12.23) and in SPSS (SPSS Inc.,
Chicago, Illinois, 1989-1999, 9.0.1), respectively.
This was necessary because Minitab could only
perform K-means cluster analysis for 2 or more
variables, therefore SBL (a single variable) was
analysed in SPSS. The regressions were fitted in S-
122
PLUS (MathSoft, Inc., Seattle, 1999, 5.1).
Fifthly, the data mining approach, Classification
and Regression Trees (CART), a technique that
generates a binary decision tree, was used to classify
the observations. In this approach, the set of data is
progressively sub-divided based on values of predictor
variables into groups that contain higher proportions
of “successes” and higher proportions of “failures”.
The relative importance of the predictor variables
is assessed in terms of how much they contribute to
successful splits into more homogeneous sub-groups.
The classification is most commonly carried out using
the Gini criterion, which always selects the split that
maximises the proportion of “successes” in one of
the groups (Petocz, 2003). Data mining techniques
are attractive because no distributional assumptions
are needed, data sets can have missing data and
analyses are less time consuming. The training data
used to create the binary decision set was the set of all
animals that have already been determined to be adult
males, immature males and mature females. SPSS
Clementine 12.0 was used for the analysis.
Finally, Minitab was also used to perform
hierarchical clustering and produce dendrograms
showing the degree of similarity of the skull data for
males, females and immature males. In general, the
conclusions reached were similar to those from the
CART analysis: it was possible to distinguish mature
males from immature males and mature females but
it was not possible to clearly distinguish immature
males from females.
Unless otherwise stated values are means quoted
+ standard errors with the number of data points in
brackets.
RESULTS
Standard body length (SBL)
SBL ranged from 157-201 cm in males (n =
33, SBL was not recorded for 4 of the adult males)
and 135-179 cm in females (n = 18). Mean lengths
were 182.9 + 2.3 (n = 33) and 149.1 + 2.5 (n= 18),
respectively. The two sample t-tests on our data
indicated that adult males were significantly larger
than adult females (Table 1). The ratio of mean female
SBL to mean male SBL was 1:1.23.
K-means cluster analysis successfully identified 2
relatively homogeneous groups from the pooled
data, i.e., cluster 1, predominantly males and cluster 2,
predominantly females (Table 2). Of the 18 females,
17 (94%) were correctly classified. Of the 33 males, 28
(85%) were correctly classified.
Proc. Linn, Sec. N.SW.13 8, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
4S9)-) o[dures OM],
q[vule]
saqqulie [NYS
(4 = W) | (Zp) (OrZ'0) | (E81) uy (ILD) | (700'0) | (Zr'0) BT [IXeUL
xed <W| 0S $000'°0 > vl g'€ 60| 066 JO pus 10119}sod 0} UoImeUD ET d
(24H < W) (S000'0 >) (08°9) | (€00°0) | (ZT0)
BIG DJUIUBSIG jSO}VOL
wml TA anno = SI a6 sl oe Yypeoiq ourURdI jSOJVOIH) ZI d
(x#IN < A) (S000'0 >) gy | (8S) | (€00°0) | (970) gn Gumeansod sedltin 10 mPEMET ME
ax <I is S000'0 > gs LY 90| 6S
(4 = W) (v07 0) (SL°9) | (L00°0) | (1r'0)
SIOSIOUI O 010U Je1eTe
xx < WW S000°0 > el 6L 91 | O88 eae ee Oe
[eqeled
(xs < A) | (EE) | (S000'0>) | (Z8°Z) (60'S) | (S000) | (6€°0)
aSsed UISI O UpRrol
J=IN | 6 680°0 a el Sis [an ONES poe Bese
xxl < W) | (ce) | (S000:0 >) (v00'0) | (170)
sossoooid jeyquovidns je Wypeor
xxd <W BE 50000 > OT 6b es ana
(xd < WD | (v0) | (S0000>) | (Z's) | 9, | (sr0r) | (€00°0) | (eT 0) (Z00°0) HOTeRSTIOS jEMOROTEATT IAST LCT
ved <W| 92 $0000 > L61 VI 60] O8Z 50 ee :
(xs < WW) | (72) | (S000'0>) | (6's) (sts) | (£000) | (S70) (Z00'0)
sassoooid yey1qio0o1d ye Bor
eet <WW| €€ 5000°0 > Gallen i 69 OT SES we a 60 a Ue ae
(4 =W) | (Zr) (8460) | (€0'0) (679) | (€00°0) | (810)
STeSeU IO SUN] 1S9}8Or
ext <I | 67 $000'0 > 6S el as LO | GLE ze UUs eee
(4 =W) (Sve 0) (Z48°9) | (Zo0'0) | (110) (100'0)
SoreUu JOTIOJUC JO IM 189891
ell IN enann=: SI i eo | pz 9€ 6 eo 119]UB JO YIPIM 1s9je9IH pq
(.A < WW) (S000'0 >) oy (97 y) | (€00°0) | (rE 0) 9 (ZL'v) | (€00'0) | (9€'0) sjeseu
xed < IN 5000°0 > g'¢ OT SCL 78 Cl | GB JO pus 1o110}sod 0} uoryjyeunH Eq
(4 =W) | (6p) (1100) | (9°72) ar (L0°7) | (700°0) | (98°0) a (ere) | (S00'0) | (88°0) ysoro
xed <WW | 0S $000'0> vI| 678I OL YE || LLNe [eqdio90 Jo s[pprur 0} uolIjeUH 7q
| cet SIN || OS $000°0 > Lo || Qk CE FL | CCIE | LE (TAD) WI3us] [eseqo[ApuoD Iq
[vs10g
at d u ye) | “aS | uvow
IZIS UBD 1usIg
—— ee SS
‘SaIMpIIOAd JUSUTIANSVOUT [[NYS JO UONALAIsep & 10J (QNOZ) [Ee 19 UOSPAEMIIC 0} AIJoY *S}9YIv.AG Ul UBAIS d1¥ (TAD) ySus] [eseqoyAp
-U09 0) dANL[AI SJUIWIIANSvAUI [NYS °
UBILIFY YINOG ovules puv seul WOAZ (WID) SYYSUa, Apog pA“pUL)S PUL (WUT) sJUdUIIANSvOUT [[NYS 10J (U 2p “A*-d “A'S ‘UvIM) sINSRIS ATvUIUINS =] 9[qQUy,
(3s9}-} a[d ues 0A}) SXaS 0A} OY) JO ULIU 94} UIIAMJOq UOSIAvduIOD pue “(smppsnd snqyisnd snjpydaz0jI47/) sjeas Any
W283
Proc. Linn. Soc. N.S.W., 131, 2010
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
(xd < WW)
xxi < W
(.d < W)
xxd < W
(,A < W)
xxd <
(xd < W)
xed <I
(9000'0)
$0000 >
(S000°0 >)
¢000'0 >
(9000°0)
$000'0 >
(8Z0'0)
$000'0 >
($000°0 >)
$000'0 >
($000°0 >)
$000°0 >
(L110)
$000°0 >
(xxd < W)
xed < W
(88€°0)
$000°0>
(0Z0'0)
$000'0>
(Z00'0)
$000°0 >
(S000'0 >)
$000°0 >
(44H < W)
xxd < W
IIUIAIJFJIG
dZIS JUBIYIUSIS
489}
(S000'0 >)
$000'0 >
(Z00'0)
£0
(S00'0)
"|
—)
~
ww
mom
a
a
o
~
loa)
a)
(9)
(Z00'0)
Sil
JSO1O [B}ISES JO IYSIO] VLZT
prose
JO WOO ye []NYS JO WYSIOH 97]
ssoooid [eIgi0c1d
JO Jopr1od purg 0} WoMpeuy $71
[EIQIOVIJUI USWILIOJ 0} UOIYJLUD p7’]
[e.19}e'7]
plosA19}d Jo puaq 0} uolseg ¢7d
Ypeoig pilose 77d
Ypeosg [eLeAyeD 17d
(10119}UR)
JOOI D1VRUIOBAZ 0} UOISeg YZ
(900'0)
A
(Z00'0)
61
\O
(99)
(6£ 01) | (Z00'0)
CCl 50
(SOIT)
vel
(Z00'0)
c0
(Z00'0)
9°0
(1000)
$000°0 >
(L701) | (100'0)
O11 €0
-} o]duies omy,
g[vuley
(6101)
CE]
EE
foe)
yjeosg oeUIOSAZIG 61d
ssoooid prous]sjsod
JO Joprog pury o} UoNpeuys) std
¢ ouIURd\sod ye oWe[ed Jo YIpeolg Lid
€ ouTURD}SOd je aye[ed Jo WIpeosg 91d
| ourlues}sod je aje]ed Jo yIpeolg Sid
elIxeu
JO JOOI ONVUIOSAZ JO WIPRdIg FId
Saqquliva [INAS
ponunuos | o[quy
Proc. Linn. Soc. N.S.W., 131, 2010
124
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
‘OO X UROUI/"A'S UOTLLIVA JO JUSIDYJI09 St ‘AO
“sgoueLieA [enboun 10} SUIMOT]e 1S9}-} o[dures OM} B OJ Poye[NI]ed OM SonyeA Jp
“UOTIOLION TUOLIOJFUOG YIM “[OAd] %] BY) Je JWROYIUSIS ,,
“UOTJNIIIOD TUOLIOJFUO YIM “JOAS] %E OY) 1 JUROIUSIS ,
(066 ‘SIOA) UONLULIOJSUeN XOD-xOg SUISN POWLIOJsuLI SPA BIeP OY} DIOJOIOYY ‘(poMmoys vyep) ploy jou prp uoNdumnsse Aplyeuttou j-syuSpNy|S
‘JYBIOY JSOIO o[eUID} UL} 1o}BOIS seM JYSIOY
JSO1O J]RUL ‘S[BUITUL OBIE] Ul “IOADMOP “SO[CUOF IOJ SJUOUIOINSLOU MJ 00} O1OM JOY} OSNed9q AT[eOT]sIeIS poULUIeXd JOU SEM (/7]) ISOIO [e}ISES FO JYSIOH &
(x24 < W)
(1 =W)
xxi < W
|r 5000'0 > vor | 81
(9Z00'0)
Iv
(L€)
8P
xed < W 8V
(4 =W) | (87)
xed < W
[iS aule:
IUIIIFIG
IZIS JUBIIUSIS
(4 = W)
xed < W OV
(4 =W) | (S72)
xx < WI 6£
—
(Z7'€)
78
$0000 >
($000'0 >)
60000 >
(100'0 >)
$0000 >
$9}-} o[durvs OM],
(100'0 >) | (OZ €y)
Al
s000'0> | O0T1
(g€0'0) | (O72) ti
S000°0 > 901
Gi
By TALON |
(61 II
) | (L00°0)
601 ST
(SL'7) | (S000) | 6
Gi | @
gs
(1dS) WSus] Apog pavpurys
ocd
On| 9E
I | as
(100'0) ee
ULy
SNOPlOuUOIO) OF sliejnsuy CLIN
SNyVoUl je O[QIPULU JO IYSIOH{ TEIN
MOI oUTURD}SOd IOMO] JO YSUST OE
A)
‘ars | ueour |
S9[QUIIvA [[NYS
(200°0) | (670) MOL Y}O0} Ie[NGIpueWl JO WSUS] 6ZINI
80 | 669 :
(z00'0) | (0L'0)
dTqQIPUeUT JO WSUS
alone [qipueul Fo YISUIT 8ZIA
Cem) a ae eeke 2 os Sisbes Seeley
ponunuos [ I[qeL
2S
Proc. Linn. Soc. N.S.W., 131, 2010
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
Table 2: Classification of skull measurements of South African fur seals using K-means clusters analy-
sis. n is the number of animals. All variables except standard body length (SBL) were standardised
(dorsal, palatal and mandibular).
Skull variables
Male
Dorsal =
Female
Sex| Gust [ste
22 (96%) a
11 (100%)
Male
24 oo 2 (8%)
Lateral
Female
Female: (lo Glee sen
Male
17 (100%)
x =m 7
10 (100%)
25 (93%)
Mandibular
16 (94%)
2 (7%) 27
17
1 (6%)
28 (85%) 5 (15%)
Standard body length
Skull variables
Absolute skull size: two sample t-tests
The two sample t-tests indicated that 30 of
the 31 mean skull variables were significantly
larger in males than in females, i.e., we reject H, in
favour of H,: [> Mima (Table 1, Fig. 1). Mean value
of breadth of brain case (D9) was not significantly
different for the genders (Table 1). The coefficient
of variation (C.V.) was larger in males, with the
following exceptions: least interorbital constriction
(D7), breadth of brain case (D9), gnathion to
anterior of foramen infraorbital (L24) and length
250
200
=
L 150
wm
a]
[42)
=
100
50
O
9) 50 100 150
Females (cm)
126
1 (6%) 17 (94%)
of mandibular tooth row (M29) (Table 1). Height
of sagittal crest (L27) was not examined statistically
because there were too many skulls with missing or
damaged sagittal crests.
Relative skull size: two sample t-tests
When skull variables were analysed relative to
condylobasal length (CBL, D1), males were found
to be significantly larger than females for 13 (43%)
variables: (1) gnathion to posterior end of nasals
(D3), (2) breadth at preorbital processes (D8), (3)
least interorbital constriction (D7), (4) breadth
at supraorbital processes (D8), (5)
greatest bicanine breadth (P12), (6)
breadth of palate at postcanine |
Fig. 1: Mean values of 31 skull var-
iables for male and female South
African fur seals. Numbers cor-
respond to skull variables listed in
Table 1 (numbers 1-9 correspond
to parameters D1 to D9, 10-23 to
P10 to P23 and 24-32 to L24 to
L32). Numbers above the dashed
line, males > females; numbers on
the line, males = females; numbers
below the line, females > males.
Minitab could only perform K-
means cluster analysis if there
was > 2 variables, therefore SBL
(a single variable) was analysed in
SPSS. SBL was not recorded for 4
of the 39 males (i.e., n = 35).
200
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
0.8
=
OF
Males (cm)
i
Oy
Orr
Females (cm)
Fig. 2: Mean values of 30 skull variables, relative to condylobasal
length, for male and female South African fur seals. Numbers corre-
spond to skull variables listed in Table 1 (numbers 1-9 correspond to
parameters D1 to D9, D10-23 to P10 to P23 and P24-32 to L24 to L32).
Numbers above the line, males > females; numbers on the line, males
= females, numbers below the line, females > males.
(P15), (7) breadth of palate at postcanine 3 (P16),
(8) calvarial breadth (P21), (9) mastoid breadth
(P22), (10) gnathion to foramen infraorbital
(124), (11) gnathion to hind border of preorbital
process (L25), (12) height of skull at bottom of
mastoid (L26) and (13) height of mandible at
meatus (M31) (Table 1, Fig. 2). Differences between
the genders were highly significant (P < 0.001); apart
from gnathion to foramen infraorbital (L24) and
height of skull at bottom of mastoid (L26), which
were significant at the 5% level (Table 1).
Breadth of brain case (D9) was significantly
different in “absolute size’ for males and females,
but ‘relative to CBL’ parameter D9/D1 for females
was larger than males (Table 1). Length of upper
postcanine row (P11) was larger in ‘absolute size’ in
males, but relative to CBL’ P11/D1 in females was
larger than in males (Table 1).
Proc. Linn. Soc. N.S.W., 131, 2010
0.6
The remaining 15 (50%)
variables were not significantly
different for the genders (Table
1). Since males were larger than
females in ‘absolute size’, this
suggested that the 15 variables
were proportionate to CBL
RESANCIlESS Oli SOx, Le, wne
ratio relative to CBL (D1) was
significantly different for the
genders.
The coefficient of variation
for values ‘relative to CBL’
was larger in males for about
1/3 rd of all variables (Table 1).
Exceptions were breadth at pre-
orbital processes (D6), least
interorbital constriction (D7),
palatal notch to incisors (P10),
breadth of zygomatic root of
maxilla (P14), breadth of palate
at postcanine 5 (P17), gnathion
to foramen infraorbital (L24),
gnathion to hind border of
preorbital process (125),
length of mandible (M28) and
length of mandibular tooth row
(M29). The coefficients of 2 of
these variables (least interorbital
constriction (D7) and length of
mandibular tooth row (M29))
were considerably larger in
females in both ‘absolute size’
and size ‘relative to CBL’ (M29/
D1 and D7/D1).
0.8
K-means cluster analysis
K-means cluster analysis successfully identified
2 relatively homogeneous groups from the pooled
data, i.e., cluster 1, predominantly males and cluster
2, predominantly females (Table 2). Classification
based on dorsal, palatal and mandibular obser-
vations was highly successful in recapturing the 2
groups. Classification based on lateral observations
was less successful.
Apart from | mandibular variable, all females
were correctly classified. The majority of males were
correctly classified with the following exceptions - 1
dorsal, 2 palatal, 2 mandibular and 7 lateral variables
were incorrectly classified as females (Table 2).
Misclassification occurred in small males only.
27)
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
Ln Conaylobasal Lengtin (mm)
GrGs Gi) Gels Gis Sith SUS) Sd Boab Bae Sd
Ln Seal Body Length (cm)
FIgue
Ln Mastoid length (mm)
44 4:5 46 4.7 4.8 4.9 5.0 5.1 5.2 5:3
Ln Seal Body Length (cm)
Linear regression
All transformed variables were regressed on
log, (SBL in cm). Three variables that best depicted
maximum discrimination between the sexes, using
regression, are given in Figs. 3, 4 and 5. These
were CBL (D1), greatest bicanine breadth (P12)
and mastoid breadth (P22). These plots (males
closed black circles, females grey squares) clearly
show pronounced sexual dimorphism in adult South
African fur seals, supporting findings of the two-
sample t-test and K-means cluster analysis.
Principal component (PC) analysis
128
Ln Greatest Bicanine Breadth (mm)
44 45 46 47 48 49 50 5.1 52 53
Ln Seal Body Length (cm)
Figs. 3, 4 & 5: Bivariate plot of: (3) log [CBL (D1)
(mm)] on log (SBL (cm)); (4) log [greatest bicanine
breadth (P12) (mm)] on log (SBL (cm)); (5) log
[mastoid breadth (P22) (mm)] on log (SBL (cm).
Circles, males. Squares, females.
The first 3 PCs accounted for most of the
variation. The first PC (PCI) can be interpreted
as a measure of overall skull size while PC2 and
PC3 define certain aspects of shape (Table 3).
Interpretations for the first 3 PCs for the 2 genders
are given in Table 4, together with the percentage of
total variation given by each PC. The variances of
corresponding PCs for the two genders do vary and
interpretations are dissimilar for most pairs of PCs.
Determining the gender of an isolated skull
It is claimed that it is often possible to make
a visual determination of the gender of an isolated
South African fur seal skull, provided the skull is
from an adult animal (Brunner, 1998ab). However,
visual identification based on morphology of the
skull alone can be misleading, e.g., young adult
males can be mistaken for larger, older females
and sex determination of a pup from examining
the skull alone would be very difficult. A more
objective procedure in determining sexes of skulls
would be desirable. In most practical situations if
the carcass was available for examination, the sex
would usually be determinable, however for many
museum specimens only the skull is available. The
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
Table 3: Principal component (PC) analysis of covariance matrix for adult male and adult female South
African fur seals, showing principal components, eigenvalues, proportions and cumulative proportions
of the first three principal components. Proportion gives the amount of the total variation that the PC
accounted for. Cumulative tally gives the amount the first PC accounted for, then the amount that the
first two PCs accounted for and finally the amount of total variation the first three PCs accounted for.
Height of sagittal crest (L27) was not examined statistically because there were few measurements for
females.
PCI | Pci | PCI PC Il
Dorsal Males (n = 23 Females (n = 10
D1 Condylobasal length -0.58 | -0.35 | -0.50 0.38
D2 Gnathion to middle of occipital crest Mil, || O05 ~ | O52 -0.32
D3 Gnathion to posterior end of nasals -0.28 0.30 -0.28 0.09
D4 Greatest width of anterior nares -0.10 0.16 0.03 0.06
D5 Greatest length of nasals -0.16 | 0.34 0.02 0.04
D6 Breadth at preorbital processes -0.19 0.30 -0.28 -0.17
D7 Least interorbital constriction -0.08 0.29 0.09 -0.37 ; -0.14
D8 Greatest breadth at supraorbital processes -0.08 | 0.49 0.38 -0.36 -0.43
D9 Breadth of brain case =0103 NN 048 -0.15 0.71
Eigenvalue 444.9 | 36.1 93.7 12.7
Proportion 0.84 0.07 0.03 0.68 0.09
Cumulative 0.84 0.91 0.94 0.68 0.91
-| Palatal Males (n = 26 Females (n = 16)
O82 NEO 349 forssHi|i0!32
P11 Length of upper postcanine row -0.13 0.10 -0.08 | -0.06 | -0.02
P12 Greatest bicanine breadth -0.01 -0.20_| -0.08 -0.19
P13 Gnathion to posterior end of maxilla -0.06 -0.24 | 0.04 0.10
P14 Breadth of zygomatic root of maxilla -0.01 -0.003 -0.03 | -0.04 0.04
P15 Breadth of palate at postcanine 1 0.03 -0.14 -0.11 | 0.08 Mh
P16 Breadth of palate at postcanine 3 -0.08 0.04 -0.08 -0.03 | 0.09 -0.24
P17 Breadth of palate at postcanine 5 -0.10 0.05 -0.14 -0.02 | 0.08 -0.24
P18 Gnathion to posterior border of postglenoid | -0.50 | -0.18 -0.06 -0.41 | -0.16 -~0.21
P19 Bizygomatic breadth -0.30 | 0.86 0.23 -0.53__ | -0.15 0.27
P20 Basion to zygomatic root -0.41 -0.11 -0.13 -0.30 | 0.13 -0.66
P21 Calvarial breadth -0.25 0.13 -0.31 -0.26 | -0.15 0.19
P22 Mastoid breadth -0.39 0.05 -0.28 -0.37 | -0.42 OW
P23 Basion to bend of pterygoid -0.13 -0.08 -0.13 -0.13 | 0.14 0.26
Eigenvalue 507.1 | 84.4 35.0 155.5 | 44.4 13.9
Proportion 0.73 | 0.12 0.05 0.62 | 0.18 0.06
Cumulative 0.73 0.85 0.90 0.62 | 0.79
Lateral Males (n = Females (n = 10)
L24 Gnathion to anterior of foramen infraorbital | 0.39 -0.56 0.73 -0.71 0.66
aa Paeon to posterior border of preorbital 0.43 0.59 0.68
L26 Height of skull at base of mastoid 0.82 0.58
27a Height of sagittal crest - - -
Eigenvalue 153.8 | 14.5 0.7 31.4
Proportion 0.91 | 0.09 0.004 0.82 | 0.16
Cumulative 0.91 1.00 0.82 | 0.98
Mandibular Males (n = 26 Females (n = 16
=O 20MMIEOSS
0965023
M31 Height of mandible at meatus 0.05 0.50
M32 Angularis to coronoideus -0.31 -0.30 | 0.14 0.66
Eigenvalue 8.0 88.5 | 27.2 9.1
Proportion 0.21 0.07
Cumulative 0.91 0.98
Proc. Linn. Soc. N.S.W., 131, 2010 129
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
Table 4: Interpretations for the first 3 principal components for the skulls parameters for adult male and adult female South African fur seals. Variables
that contributed predominantly to size and/or shapes, i.e. variables with loadings = 0.36 (absolute value) were used in the covariance matrix. Only 2
principal components were considered for the analysis of lateral components because component 3 was 2% or less of total variation.
Dorsal
Component | (male 84%, female 68%
CBL (D1), breadth at preorbital processes (D6), least interorbital constriction
(D7) and greatest breadth at supraorbital processes (D8) measures overall size.
Component 2 (male 7%, female 13%
Contrasts greatest breadth at supraorbital processes (D8) with CBL (D1) and Contrasts CBL (D1) with gnathion to posterior end of nasals (D3), greatest
breadth of brain case (D9). breadth at supraorbital processes (D8) and breadth of brain case (D9).
Component 3 (male 3%, female 9%)
Contrasts CBL (D1) with gnathion to middle of occipital crest (D2), greatest Contrasts greatest breadth at supraorbital processes (D8) with CBL (D1) and
CBL (D1) and g~athion to middle of occipital crest (D2) measure overall size.
breath at supraorbital processes (D8) and breath of brain case (D9). breadth of brain case (D9).
Palatal
0
Gnathion to posterior border of postglenoid process (P18), bizygomatic breath
P19) and mastoid breadth (P22) measure overall size.
Component 2 (male 12%, female 18%)
Gnathion to posterior border of postglenoid process (P18), basion to zygomatic
root (P20) and mastoid breadth (P22) measure overall size.
Female
Contrasts palatal notch to incisors (P10) with mastoid breadth (P22).
Component 3 (male 5%, female 6%)
Bizygomatic breadth (P19) dominates.
Bastion to zygomatic root (P20) dominates.
Lateral (only 2 PCs considered)
Component | (male 91%, female 82%)
Height of skull at base of mastoid (L26), gnathion to posterior border of
preorbital process (L25) and gnathion to anterior of foramen infraorbital (L24) Height of skull at base of mastoid process (L26) measures overall size.
measure overall size.
Palatal notch to incisors (P10) dominates
Component 2 (male 9%, female 16%
coronoideus (M32) measure overall size. overall size.
Component 2 (male 8%, female 21%) _
Contrasts height of mandible at meatus (M31) and angularis to coronoideus
(M32) with others [length of mandible (M28), length of mandibular tooth row Length of mandibular tooth row (M29) dominates.
(M29), length of lower postcanine row (M30)].
Component 3 (male 5%, female 7%)
Contrasts length of mandible (M28) with length of mandibular tooth row (M29) Contrasts length of mandible (M28) and length of lower postcanine row (M30)
and height of mandible at meatus (M31). with height of mandible at meatus (M31) and angularis to coronoideus (M30).
Contrasts height of skull at base of mastoid (L26) with gnathion to anterior of Contrasts height of skull at base of mastoid (L26) with gnathion to anterior of
foramen infraorbital (L24) and gnathion to posterior border of preorbital process | foramen infraorbital (L24) and gnathion to posterior border of preorbital process
(L25). (L25).
Mandibular
omponen male 84%, fema 0%
Length of mandible (M28), height of mandible at meatus (M31) and angularis to | Length of mandible (M28) and height of mandible at meatus (M31) measure
Proc. Linn. Soc. N.S.W., 131, 2010
130
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
sex of tagged individuals would nearly always be
known, as it would have been recorded when they
were tagged.
We have focused on trying to develop a method
for making an objective determination of sex based
on only skull material. Aging untagged specimens
from dentition (counting the growth layer groups
in the upper canine) is an important component of
making an objective sex determination.
The skull of an adult male > 10 y is larger (CBL
= 248 mm; mastoid breadth > 134 mm) and more
robust than the skull of a similar aged female. In adult
males, bony deposits occur throughout the parietal
region of the skull, which become more prominent
with increasing age (Rand, 1949ab; Stewardson et
al., 2008; present study). Mean size of male sexually
dimorphic traits, according to age (y), have been
summarised elsewhere (Stewardson et al., 2008,
2009).
Classification and Regression Tree using 3
levels (58 animals)
Fig. 6 shows an animal is classified as being
an immature male if 125<=73.7, P12<=35.85
and P16<=17.24 or if 125<=73.7, P12>35.85
and M32<=50.5 or if 125>73.7, P12<=45.1 and
D5<=41.65. An animal is classified as being
a mature female if 125<73.7, P12<=35.85 and
P16>17.25 or if 124<=73.7, P12>35.85 and
M32>50.5. An animal is classified as being
a mature male if 125>73.7 and P12>45.1 or if
125>73.7, P12<=45.1 and D5>41.65. This rule
correctly classifies 94.82% of the animals.
Three immature males are misclassified as
being a mature female (15% of all immature
males). All mature females are correctly class-
ified as being mature females, and all mature
males are correctly classified as being mature
females. Fig. 6 includes a prediction matrix to
summarise the classification of the animals.
Hierarchical Cluster Analysis of skull parameters to
produce a dendrogram (30 animals)
Cluster analysis was performed on_ thirty
individuals where data on all variables were available,
not counting SBL and sagittal crest height (L27). The
observations were clustered using complete linkage
(furthest neighbour) and Euclidean distance on all
variables excluding SBL and L27. The four immature
males lacking SBL data and hence for which there
was some doubt about their actual sex (PEM 2006,
2009, 2010 & 2014) were excluded from the analysis.
Cutting the dendrogram (Fig. 7) at a similarity level
of 66.67 (or distance of 90) produces four clusters.
Proc. Linn. Soc. N.S.W., 131, 2010
The first cluster contains 2 males, 6 immature males
and 2 females: PEM 975-M, PEM 2048-M, PEM
1014-F, PEM 1138-F, PEM 2046-IM, MCM 4577-
IM, MCM 5133-IM, PEM 2050-IM, PEM 2052-IM,
and PEM 2081-IM. The second cluster contains all
males (10/10): PEM 1453-M, PEM 1892-M, PEM
2049-M, PEM 2051-M, PEM 2054-M, PEM 2087-
M, PEM 2140-M, PEM 2141-M, PEM 2143-M, and
PEM 2151-M. The third cluster contains 4 immature
males and 3 females: PEM 2084-F, MCM 4578-
F, MCM 5154-F, MCM 4595-IM, MCM 4996-IM,
MCM 5002-IM, and MCM 5135-IM. The fourth
cluster contains one female and 2 immature males:
MCM 4994-F, MCM 4989-IM and MCM 5145-IM.
Inclusion in the dendrogram of SBL data did not
improve the ability to distinguish between immature
males and females. Thus using cluster analysis it
is easily possible to distinguish mature males from
immature males and females but it is not possible to
separate immature males from females.
DISCUSSION
Possible bias
Several factors must be taken into
consideration when interpreting the data. Firstly,
the sample size is small; in particular only 6 of
the 14 females were aged. Secondly, there may be
an over representation of either larger or smaller
individuals in the data set which may possibly bias
the results. Thirdly, although identical variables
were taken from PEM and MCM animals, PEM
variables were recorded by the first author, whereas
MC™M variables were recorded by the third author,
introducing possible inter-observer error. However,
the most likely source of bias is that some of the
museum specimens identified as immature males
may have been incorrectly sexed, especially if only
the skull had been collected and the carcass had
not been inspected properly, was badly decayed or
was not available for examination. The results of
the Classification and Regression Tree (Fig.
6) and the Cluster Analysis dendrogram (Fig. 7)
emphasize that caution should be taken about the
common claim that male and female skulls can
be distinguished by visual inspection (Brunner
1998ab). The Classification and Regression
Tree analysis was the more successful in correctly
identifying the sex of the skulls. The cladistic
dendrogram method had no difficulty in recognising
mature male skulls but female and immature male
skulls cannot be objectively separated from one
another.
131
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
<= 35.850
Node 3
Category % n
"1.000
® 2.000
@ 3.000
Total
0000 0
83.333 10
16.667 2
20.690 12
<= 73.700
Node 1
Category % fn
1.000
0.000 0
43.750 14
56.250 18
55.172 32
=2000
@ 3.000
Total
Sex
Node 0
Category % ‘
® 1.000
= 2.000
@ 3.000
Total
» 35,850
= 2.000
=3.000
Total
Node 4
Calegoy % n
= 1.000
0.000 0
20.000 4
80.000 16
34.483 20
41.379 24}:
24.138 14]:
34.483 20}!
100.000 58};
<= 45100
Node 5
Categoy % on
= 1.000
= 2000
@3000
Total
66.667
» 73.700
Node 2
Category % n
"1.000
92.308 24
0.000 0
7.692 2
44.828 26
=2000
@3000
Total
33.333
0.000
5.172
» 45.100
Node 6
Category % n
m32
<= 17.250 > 17.250 <= 50.500 > 50.500 <= 41.650
Node 12
Category % on
100.000 1
0.000 0
0.000 0
1,724 1
Node 11
Categoy % on
= 1.000 0.000 0
Node 10
Category on
= 1.000
Node @
Category % fn
= 1.000 0.000 0
@2000 0.000 0
@3000 10000013
Total 22.414 13
Node 8
Category % fi)
© 1.000 0.000 0
@2000 10000010
@ 3.000 0000 0
Total 17.241 10
Node 7
Category % n
42.857 3
12069 7
100.000 2
3448 2 Total
Total
Prediction Matrix for 3-level Classification (n and % )
Predicted Adult Male Predicted Female (2)
(
1)
Fig. 6: Classification and Regressions Tree (CART) using three levels of skull data sets of adult male (M),
immature male (IM) and female (F) South African fur seals (Total n = 58). A table is included to indicate
successful and unsuccessful determinations of sex (M/F) and male reproductive status (IM/M). All the
adult males (n = 24) were successfully identified as adult males. Three (3) immature males or 15% of
the total (n = 20) were incorrectly classified as females but all the known females (n = 14) were correctly
identified as females.
Predicted Immature
Male (3)
Immature Male (3)
size (in multidimensional space). Gnathion to
middle of occipital crest and basion to zygomatic
root were predominant in males but not in females.
Bizygomatic breadth was predominant in females -
but not in males.
Principal component analysis: skull size and
shape
For both genders, CBL, mastoid breadth,
height of skull at base of mastoid, gnathion to
posterior border of postglenoid process and length
of mandible contributed the most to overall skull
132 Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
b abd ob % ab F cb ab xb
eS Sg OP ere y
rag My CHOON ROX LON
FPG COL VO LO OSCE SE
Observations
Fig. 7: Cladistic dendrogram based on complete sets of skull data for adult male (M), immature male
(IM) and female (F) South African fur seals (Total n = 30). At the 66.67% similarity level the dendrogram
divides into four groups or clades. One clade (#2) at the centre consists entirely of mature males (10/10)
but the other three groups consist of two mature males (M), and a mixture of immature males (IM) and
females (F). Clade (#1) consists of 2 females, 2 males and 6 immature males, clade (#3) consists of 3 fe-
males and 4 immature males and clade (#4) consists of 1 female and 2 immature males.
Predominant variables contributing to shape in
both genders were CBL, breadth at supraorbital
processes, breadth of brain case, palatal notch to
incisors, gnathion to anterior of foramen infraorbital,
gnathion to posterior border of preorbital process,
height of skull at base of mastoid, length of mandible,
length of mandibular tooth row, length of lower
postcanine row, height of mandible at meatus and
angularis to coronoideus (see figures of South African
fur seal skulls in Stewardson et al., 2008).
Bizygomatic breadth contributed predominantly
to skull shape in males but not in females. Gnathion
to posterior end of nasals, basion to zygomatic root
and mastoid breadth contributed predominantly to
skull shape in females but not in males.
These findings indicate that the underlying data
structure for males and females was different. Dif-
ferences occurred in the combination of predom-
inant variables, and in their magnitude and sign.
Proc. Linn. Soc. N.S.W., 131, 2010
General pattern of growth
Although male South African fur seals are slightly
heavier than females (4.5 vs. 6.4 kg) at birth, growth
patterns for the genders are reportedly similar up until
puberty (Warneke and Shaughnessy, 1985). Males
attain puberty between 3 and 4 y (Rand 1949b;
Wareke and Shaughnessy, 1985; Stewardson et al.,
1998) and females between 3 and 5 y (Rand 1949a;
Warneke and Shaughnessy, 1985; Guinet et al., 1998,
J.H.M David, pers. comm.).
Although males are sexually mature at an early
age, they are physically unable to hold a harem until
much later. Full reproductive status (social maturity)
is deferred until full size and competitive vigour are
developed. Males normally do not reach breeding or
“beachmaster” status until about 10 y (Rand, 1949b;
Stewardson et al., 1998). Some never attain breeding
status. Females approximate adult size at about 5 y of
age, while males attain adult size between 8 and 10
y (Rand, 1949a; Stewardson 2001; Stewardson et al.,
2008, 2009). Adult males may weigh up to 353 kg
133
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
(mean, 250 kg), while females may weigh up to 122 kg
(mean, 58 kg) (David 1987; Guinet et al., 1998; J.H.M
David, pers. comm.).
Redigitising the Australian fur seal data from
Arnould and Warneke (2002), as described previously
in our study of body size in male Australian and
South African fur seals (Stewardson et al., 2009), it
was possible to estimate the SBL of adult (>135 cm)
female Australian fur seals to be 157+ 0.758 (n= 144)
cm. A two-sample t-test shows that Australian female
fur seals were significantly larger than South African
female fur seals (p < 0.001) but the overall difference
is small (7.9 + 2.6 cm). Guinet et al. (1998) based on
adult females shot at a breeding colony in Namibia
found the mean SBL of female South African fur
seals to be 147 + 0.56 cm (n = 157), which is not
significantly different to that calculated in the present
study (Appendix 3: 149 + 2.49 cm, n= 18) . A two-
sample t-test using their data, with its much larger
sample size, leads to the same conclusion that female
South African fur seals are slightly smaller than their
Australian counterparts. These results are similar to
the finding in male South African vs. Australian fur
seals that the South African form of Arctocephalus
pusillus is slightly smaller than the Australian variety
(Stewardson et al., 2009). Overall then, both male
and female South African fur seals are smaller than in
the case of the Australian fur seal.
Studies of increase in SBL vs. age consistently
show monophasic post-weaning growth patterns
with different growth kinetics for each sex in the
South African fur seal (Stewardson et al., 1998,
2008, 2009), Australian fur seal (Arnould and
Warneke, 2002; Brunner et al., 2004; Stewardson
et al., 2008, 2009) and other polygynous
breeding pinnipeds which exhibit pronounced size
dimorphism, e.g., Antarctic fur seal (A. gazella) and
Southern fur seal (A. tropicalis) (Daneri et al., 2005),
New Zealand fur seal (A. australis forsteri) (Brunner,
1998b; Brunner et al., 2004; McKenzie et al., 2007),
Northern fur seals (Callorhinus ursinus) (McLaren,
1993) and the Steller sea lion (Ewmetopias jubatus),
based on several hundred individuals (Winship et al.,
2001).
Development of the skull in male South African
fur seals exhibits monophasic growth in some
variables and biphasic growth in others (Stewardson
et al., 2008, 2009). In males, biphasic growth in skull
parameters is associated with reaching an age of
about 8 to 10 y when some males attain full-breeding
status (Stewardson et al., 2008). Similar growth
patterns have been reported in the skulls of male
New Zealand fur seals (Brunner, 1998ab; Brunner et
al., 2004). There does not appear to be sufficient
134
size/age data available to make statements about
the growth dynamics of the female skull of any of
the fur seal species.
Variation among adult males
The coefficient of variation for most skull
variables was larger in males than in females
(Stewardson et al., 2008; present study). Variability
in adult males at least partly reflects differences
in social status. Differences in physical appearance
will be most noticeable before and during the breeding
season when breeding bulls build up their body
reserves. The specimens used in the present series
of studies of South African fur seals (A. pusillus
pusillus) (Stewardson et al., 2008, 2009) were
based on fur seals collected from feeding areas
on the eastern coast of South Africa rather than
from breeding colonies and so would consist of
a mixture of breeding and non-breeding animals.
Data available on Australian fur seal (A. pusillus
doriferus) are based on animals collected from
breeding colonies (Arnould and Warneke, 2002;
Brunner et al., 2004).
Loci of sexual dimorphism
Dorsal
Males were significantly larger than females
‘relative to CBL’ in four of the nine dorsal variables
(gnathion to posterior end of nasals (D3), breadth
at preorbital processes (D6), least interorbital
constriction (D7), breadth at supraorbital processes
(D8)). In both genders, these variables form
part of the splanchnocranium (gnathion to
posterior end of nasals (D3)) and the frontal region
(least interorbital constriction (D7) and breadth at
supraorbital processes (D8)), and are associated
with respiration/vocalisation (gnathion to posterior
end of nasals (D3)) and feeding (breadth at
supraorbital processes (D8)).
Inmales, at least two of these variables have obvious
functional significance with respect to territorial
acquisition and defence. Least interorbital
constriction (D7) and breadth at supraorbital
processes (D8) contribute to the structural strength of
the skull, and shield the animal against blows to the
head (especially the eyes) during combat with rival
males. They also increase the width of the face of the
seal, making it appear more intimidating to its rivals.
Palatal
Males were significantly larger than females
‘relative to CBL’ in five of the 14 palatal variables
(greatest bicanine breadth (P12), breadth of palate ©
at postcanine | (P15) and postcanine 3 (P16),
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
calvarial breadth (P21) and mastoid breadth
(P22)). In both genders, greatest bicanine breadth
(P12), breadth of palate at postcanine 1 (P15) and
postcanine 3 (P16), form part of the palatal region and
are like other parameters from that part of the
skull (greatest bicanine breadth (P12), breadth of
palate at postcanine | (P15) and postcanine 3 (P15))
are associated with feeding and respiration /
vocalisation (greatest bicanine breadth). Calvarial
breadth (P21) and mastoid breadth (P22) form
part of the basicranium and are associated
primarily with auditory function (calvarial breadth
(P21), mastoid breadth (P22)).
Enlargement of the canines (greatest
bicanine breadth (P12)) enables males to inflict a
potentially lethal bite during combat. The rostrum
is broad (palatal breadth at postcanine 1 (P15)
and postcanine 3 (P16)), accommodating the large
canines. Enlargement of calvarial breadth (P21) and
mastoid breadth (P22) increases intimidating size of
the face and increases the structural strength of the
skull (large head size/ mass).
Lateral
Males were significantly larger than females
‘relative to CBL’ in all lateral variables; that is,
gnathion to anterior of foramen infraorbital (L24),
gnathion to hind border of preorbital process (25)
and height of skull at bottom of mastoid (26). In both
genders, gnathion to foramen infraorbital (L25)
and gnathion to hind border of preorbital process
(L25) form part of the splanchnocranium and
are associated with respiration/ vocalisation.
Enlargement of skull height and facial length in
males increases the overall head size.
Mandible
Males were significantly larger than females
‘relative to CBL’ in only one mandibular variable
(height of mandible at meatus, M31). This variable
is associated with auditory function and feeding
in both genders (Stewardson et al., 2008).
Enlargement of this variable in males increases gape
and provides a larger surface area for muscle
(masseter and temporalis) attachment. Large jaws
and jaw muscles are advantageous in territorial
combat.
Significance of the dimorphism
In male South African fur seals, there appears to
be strong selection pressure for the development of
certain morphological traits associated with fighting
ability and body size and mass. It is important to
note that beachmasters spend much of their time
Proc. Linn. Soc. N.S.W., 131, 2010
vocalising and intimidating rivals by displays which
emphasise their size and the likely consequences
of a rival attempting to challenge them rather than
actual fighting (Rand, 1967; Stirling and Warneke,
1971; Miller, 1991). In male South African fur seals,
selection pressure appears to favour large body
mass. Stewardson et al. (2008, 2009) showed that
males (mean, 183 cm) were significantly larger in
standard body length than females (mean, 149
cm). Thus, on the mass/length cubed rule one
would expect a male to weigh about 2 times that
of an average female. Relative differences in body
mass are much higher: large males in breeding
condition may be 4-5 times heavier (average
about 250 kg) than adult females, which average
about 58 kg (David, 1989; Guinet et al., 1998; J.H.M
David, pers. comm.). Large males have an advantage
over their smaller rivals in gaining high social rank
through vocalisation, intimidating display and
fighting (Stirling and Warneke, 1971; Miller, 1991).
Furthermore, large males in breeding condition have
a well developed fat store. This thick blubber layer
enables males to remain resident on territory for long
periods (up to 40 days) without feeding and provides
protection as well (Peterson, 1968; Alexander et
al., 1979; McCann, 1981; Campagna and Le Boeuf,
1988; Boness, 1991). As in most seals, 1f for any
reason a male abandons his territory, it will quickly
be occupied by a rival male and the usurper will most
likely have to be removed by actual combat (Rand,
1967; Le Boeuf, 1974; Miller, 1974; McCann, 1981;
Campagna and Le Boeuf, 1988). There is a high risk of
injury and/or failure in attempting to regain breeding
territory.
Selection pressure also appears to favour the
development of certain skull traits that appear to be
associated with potential and actual fighting ability.
In the present study, traits which are significantly
larger in males appear to be associated with bite
force (e.g., broad canines, increased surface area
for muscle attachment, large gape), large head size/
mass (e.g., increased mastoid and calvarial breadth)
and/or structural strength of the skull (protection
against damage from direct blows to the head during
combat).
Sexual dimorphism of the skull in southern fur
seals has also been reported for the Australian
and New Zealand fur seals (Australian fur seal, A.
pusillus doriferus and New Zealand fur seal, A. australis
forsteri) (Brunner, 1998ab). As with the South African
fur seal, sexually dimorphic traits are mainly those
characteristics that increase the ability of males to
acquire and defend territory in the short breeding
season whether by simply visually and vocally
135
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
intimidating potential opponents or by actual combat
(Bartholomew, 1970; Stewardson et al., 1998).
CONCLUSIONS
Information presented in the study demonstrates
that there is pronounced sexual dimorphism in adult
South African fur seals with respect to body length,
body mass, skull size and skull shape. Male South
African fur seals were significantly larger than
females in SBL, and 43% of skull variables were
found to be significantly larger in males relative
to CBL. These variables were associated with
fighting ability, e.g., large head size/mass, increased
structural strength of the skull and/or increased bite
capacity. Principal component analysis showed that
the underlying data structure for males and females
was different, and that most variation between the
sexes was expressed in overall skull size rather than
shape. This makes it generally easy to distinguish
mature male and female skulls but problematic to
distinguish skulls from sub-adult males from adult
females. Condylobasal length (CBL or D1), height
of skull at bottom of mastoid (L26) and length
of mandible (M28) contributed considerably
to overall size, with gnathion to middle of
occipital crest (D2) predominating in males only.
Classification and Regression Tree analysis
and cluster analysis dendrograms were both very
successful for distinguishing mature male skulls from
immature male and female skulls but Classification
and Regression Tree was better than cluster analysis
in distinguishing immature male from female skulls.
The material used in the present study was from a
feeding, not breeding area: it would be interesting
to attempt to determine whether breeding bulls
constitute an identifiable subset of the total adult male
population some of which never breed.
ACKNOWLEDGEMENTS
We wish to express our sincere appreciation to the following
persons and organisations for assistance with this study: Dr
V. Cockcroft (Port Elizabeth Museum), Dr J. Hanks (WWF-
South Africa) and Prof. A. Cockburn (Australian National
University) for financial and logistic support; Mr B. Rose
(Oosterlig Visserye, Port Elizabeth) who enabled us to
collect seals from his commercial fishing vessels; Dr G. Ross
(formerly Port Elizabeth Museum) and Dr V. Cockcroft for
the use of PEM skulls collected before April 1992 (n = 16
skulls); Dr J.H.M David (MCM) for the use of MCM skulls
of known-age; Mr H. Oosthuizen for assistance with aging
techniques; Mr S. Swanson (MCM) for assistance with data
136
extraction and measurement of MCM specimens; Mr N.
Minch (Australian National University) for photographic
editing; Dr C. Groves (Australian National University)
for his constructive comments on an earlier draft of this
manuscript. This paper is based upon a PhD study by C.L
Stewardson compiled on behalf of the World Wild Fund
For Nature — South Africa (project ZA-348, part 4) and
submitted to the Australian National University in 2001.
REFERENCES
Alexander, R.D., Hoogland, J.L., Howard, R.D., Noonan,
K.M. and Sherman, P.W. (1979). Sexual dimorphism
and breeding systems in pinnipeds, ungulates,
primates and humans. In ‘Evolutionary biology and
human social behaviour ’ (Eds. Chagnon, N.A. and
Irons, W.). Duxbury Press Publ., North Scituate,
Mass., USA.
Arnould, J.P.Y. and Warneke, R.M. (2002) Growth and
condition in Australian fur seals (Arctocephalus
pusillus doriferus) (Carnivora:Pinnipedia). Australian
Journal of Zoology 50, 53-66.
Bartholomew, G.A. (1970). A model for the evolution of
pinniped polygyny. Evolution 24, 546-559.
Boness, D.J. (1991). Determinants of mating systems
in the Otariidae (Pinnipedia). In ‘Behaviour of
pinnipeds ’ (Ed. Renouf, D.), pp. 1-65. Chapman and
Hall Publ., London.
Bonner, W.N. (1968). The fur seal of South Georgia.
British Antarctic Survey Scientific Reports 56, 1-81.
Bonner, W.N. (1981). Southern fur seals Arctocephalus
(Geoffroy Saint-Hilaire and Cuvier, 1826). In
‘Handbook of Marine Mammals, vol. 1: The walrus,
sea lions, fur seals and sea otter’ (Eds. Ridgway, S.H.
and Harrison, R.J.), pp. 161-208. Academic Press
Publ., London.
Brunner, S. (1998a). Skull development and growth in
the southern fur seals Arctocephalus forsteri and A.
pusillus doriferus (Carnivora: Otariidae). Australian
Journal of Zoology 46, 43-66.
Brunner, S. (1998b). Cranial morphometrics of the
southern fur seals Arctocephalus forsteri and A.
pusillus (Carnivora: Otariidae). Australian Journal of
Zoology 46, 67-108.
Brunner, S., Shaughnessy, P.D. and Bryden, M.M. (2002).
Geographic variation in skull characters of fur seals
and sea lions (family Otariidae). Australian Journal
of Zoology 50, 415-438.
Brunner, S., Bryden, M.M. and Shaughnessy, P.D. (2004).
Cranial ontogeny of otariid seals. Systematics and
Biodiversity 2, 83-110.
Bryden, M.M. (1972). Growth and development of
marine mammals. In ‘Functional anatomy of marine
mammals’ (Ed. Harrison, R.J.), vol. 1, pp. 58-60.
(Academic Press Publ., London, New York).
Campagna C. and Le Boeuf, B.J. (1988). Reproductive
behaviour of southern sea lions. Behaviour 104, 233-
261.
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
Cochran, W.G. (1977). Sampling techniques. 3rd Ed. (John
Wiley and Sons Publ., New York).
Committee on Marine Mammals (1967). Standard
variables of seals. Journal of Mammalogy 48, 459-
462.
Daneri, G.A., Esponda, C.M.G., de Santis, L.J.M. and
Pla, L. (2005). Skull morphometrics of adult male
Antarctic fur seal, Arctocephalus gazella, and the
South American fur seal A. australis. Iheringia Serie
Zoologie, Porto Alegre 95, 261-267.
David, J.H.M. (1987). Diet of the South African fur seal
(1974-1985) and an assessment of competition
with fisheries in southern Africa. In ‘The Benguela
and comparable ecosystems’ (Eds. Payne, A.L.L.,
Gulland, J.A. and Brink, K.H.). South African
Journal of Marine Science 5, 693-713.
Guinet, C. Roux, J.P., Bonnet, M. and Mison, V. (1998).
Effect of body size, body mass, and body condition
on reproduction of female South African fur seals
(Arctocephalus pusillus) in Namibia. Canadian
Journal of Zoology 76, 1418-1424.
Harrison, R.J., Hubbard, R.C., Peterson, R.S., Rice, C.E.
and Schusterman, R.J. (1968). The behaviour and
physiology of pinnipeds. (Appleton-Century-Crofts
Publ., New York).
Jackson, J.E. (1991). A User's Guide to Principal
Components. (John Wiley and Sons Publ., Hoboken,
NJ, USA).
Jolliffe, I.T. (1986). Principal Component Analysis. 2nd
ed. (Springer-Verlag Publ., New York).
Johnson, R.A. and Wichern, D. (1992). Applied
multivariate statistical analysis. 3rd ed. (Prentice
Hall Publ., Englewood Cliffs, NJ, USA).
King, J.E. (1969). The identity of the fur seals of Australia.
Australian Journal of Zoology 17, 841-853.
Laws, R.M. (1953). The elephant seal (Mirounga
leonina Linn.). 1. Growth and age. Falkland Islands
Dependencies Survey Scientific Reports 8, 1-62.
Le Boeuf, B.J. (1974). Male-male competition and
reproductive success in elephant seals. American
Zoology 14, 163-176.
McCann, T-.S. (1981). Aggression and sexual activity of
male southern elephant seals, Mirounga leonina.
Journal of Zoology (London) 195, 295-310.
McKenzie, J., Page, B., Goldsworthy, S.D. and Hindell,
M.A. (2007). Growth strategies of New Zealand fur
seals in southern Australia. Journal of Zoology 272,
377-389.
McLaren, I.A. (1993). Growth in pinnipeds. Biological
Review 79, 1-79.
Miller, E.H. (1974). Social behaviour between adult male
and female New Zealand fur seals Arctocephalus
forsteri (Lesson) during the breeding season.
Australian Journal of Zoology 22, 155-173.
Miller, E.H. (1975). Annual cycle of fur seals,
Arctocephalus forsteri (Lesson) on the Open Bay
Islands, New Zealand. Pacific Science 29, 139-152.
Miller, E.H. (1991). Communication in pinnipeds, with
special reference to non-acoustic signalling. In
Proc. Linn. Soc. N.S.W., 131, 2010
‘Behaviour of pinnipeds (Ed. Renouf, D.), pp. 128-
235. (Chapman and Hall Publ., London, U.K.).
Myers, R.H. (1990). Classical and modern regression with
applications. 2nd Ed. (PWS-Kent Publ., Boston, MS,
USA).
Payne, M.R. (1978). Population size and age
determination in the Antarctic Fur seal Arctocephalus
gazella. Mammal Review 8, 67-73.
Payne, M.R. (1979). Growth in the Antarctic fur seal
Arctocephalus gazella. Journal of Zoology (London)
187, 1-20.
Peterson, R.S. (1968). Social behaviour in pinnipeds
with particular reference to the northern fur seal. In
‘The behaviour and physiology of pinnipeds’ (Eds.
Harrison, R.J., Hubbard, R.C., Peterson, R.S., Rice,
C.E. and Schusterman, R.J.), pp. 3-53. (Appleton-
Century-Crofts Publ., New York, NY, USA).
Petocz, P. (2003). Analysing and Interpreting Information
from Questionnaires using Data Mining and Logistic
Regression. In “EMAC 2003 Proceedings: Sixth
Engineering Mathematics Applications Conference,
Sydney NSW, Australia, July 2003 (Eds. May,
R.L. and Blyth, W.F.), pp. 199-210. (Engineering
& Mathematics Applications Group ANZIAM,
Melbourne VIC, Australia).
Ralls, K. (1977). Sexual dimorphism in mammals;
avian models and unanswered questions. American
Naturalist 111, 917-938.
Rand, R.W. (1949a). Studies on the Cape fur seal
Arctocephalus pusillus pusillus 1. Age grouping
in the female. Progress report submitted June
1949, Government Guano Islands Administration,
Department of Agriculture, Union of South Africa.
Rand R.W. (1949b). Studies on the Cape fur seal
Arctocephalus pusillus pusillus 3. Age grouping
in the male. Progress report submitted November
1949, Government Guano Islands Administration,
Department of Agriculture, Union of South Africa.
Rand, R.W. (1967). The Cape fur seal Arctocephalus
pusillus pusillus 3. General behaviour on land and at
sea. Sea Fisheries Research Institute Investigational
Report, South Africa 60, 1-39.
Shaughnessy, P.D. and Ross, G.J.B. (1980). Records of
the Subantarctic fur seal (Arctocephalus tropicalis)
from South Africa with notes on its biology and some
observations on captive animals. Annals of the South
African Museum 82, 71-89.
Stewardson, C.L. (2001). “Biology and conservation of
the Cape (South African) fur seal Arctocephalus
pusillus pusillus (Pinnipedia: Otariidae) from the
Eastern Cape Coast of South Africa”, Thesis (Ph.D.),
Australian National University. http://thesis.anu.edu.
au/public/adt-ANU20030124.162757/index.html
Stewardson, C.L., Bester, M.N. and Oosthuizen, W.H.
(1998). Reproduction in the male Cape fur seal
Arctocephalus pusillus pusillus: age at puberty
and annual cycle of the testis. Journal of Zoology
(London) 246, 63-74.
137
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie,
R.J. (2008). Age determination and growth in the
male South African Fur Seal Arctocephalus pusillus
pusillus (Pinnipedia: Otariidae) based upon skull
material. Procedings of the Linnean Society of New
South Wales 129, 207-252.
Stewardson, C.L., Prvan, T., Meyer, M. and Ritchie, R.J.
(2009). Age determination and growth in the male
South African fur seal Arctocephalus pusillus pusillus
(Pinnipedia: Otariidae) based on external body
measurements. Proceedings of the Linnean Society of
New South Wales 130, 219-244.
Stirling, I. (1970). Observations on the behaviour of the
New Zealand fur seal, (Arctocephalus forsteri).
Journal of Mammalogy 51, 766-778.
Stirling, I. (1983). The evolution of mating systems in
pinnipeds. In “Recent Advances in the study of
mammalian behaviour’ (Eds. Eisenberg, J.F. and
Kleiman, D.G.), pp. 489-527. Special publication No
7, American Society of Mammalogists.
Stirling, I. and Warneke, R.M. (1971). Implications of a
comparison of the airborne vocalisations and some
aspects of the behaviour of the two Australian fur
seals, Arctocephalus spp., on the evolution and
present taxonomy of the genus. Australian Journal of
Zoology 19, 227-241.
Trillmich, F. (1984) Natural history of the Galapagos fur
seal (Arctocephalus galapagoerris, Heller). In “Key
environments — Galapagos '(Ed. Perry, R.), pp. 215-
223. (Pergamon Press Publ., Oxford, U.K.).
Warneke, R.M. and Shaughnessy, P.D. (1985).
Arctocephalus pusillus pusillus, the South African
and Australian fur seal: taxonomy, evolution,
biogeography, and life history. In “Studies of Sea
Mammals in South Latitudes’ (Eds. Ling, J.K.
and Bryden, M.M.), pp. 53-77. Proceedings of a
symposium of the 52nd ANZAAS Congress in
Sydney, May 1985. (South Australian Museum,
Adelaide, Australia).
Weisberg, S. (1985). Applied linear regression. 2nd ed.
(John Wiley and Sons Publ., New York, NY, USA).
Wickens, P.A. (1993). Life expectancy of fur seals with
special reference to the South African (Cape) fur seal.
South African Journal of Wildlife Research 23, 101-
106.
Winship, A.J., Trites, A.W. and Calkins, D.G. (2001).
Growth in body size of the Steller sea lion
(Eumetopias jubatus). Journal of Mammalogy 82,
500-519.
APPENDIX 1
Museum ascension numbers of male South African Fur seal specimens used in the present study. The data
set of skull and body measurements on these specimens has been published previously in Stewardson et al.
(2008). PEM stands for Post Elizabeth Museum (Port Elizabeth, South Africa), MCM stands for Marine and
Coastal Management (Cape Town, South Africa).
The ascension numbers of the 39 adult male animals used in the present study were:
MCM 1809, MCM 4597, MCM 4992, PEM 898, PEM 951, PEM 958, PEM 975, PEM 1453, PEM 1507,
PEM 1560, PEM 1587, PEM 1698, PEM 1868, PEM 1877, PEM 1879, PEM 1882, PEM 1890, PEM 1892,
PEM 1895, PEM 2004, PEM 2007, PEM 2013, PEM 2036, PEM 2048, PEM 2049, PEM 2051, PEM 2052,
PEM 2054, PEM 2082, PEM 2081, PEM 2087, PEM 2132, PEM 2140, PEM 2141, PEM 2143, PEM 2151,
PEM 2248, PEM 2252, PEM 2258.
The skulls classed as immature (subadult) males (n = 24) were:
MCM 2763, MCM 2795, MCM 3582, MCM 3586, MCM 3587, MCM 3636, MCM 4365, MCM 4388, MCM
4577, MCM 4595, , MCM 4996, MCM 5002, MCM 5133, MCM 5135, MCM 5136, PEM 1704, PEM 1891,
PEM 2006, PEM 2009, PEM 2010, PEM 2014, PEM 2046, PEM 2050, PEM 2053.
There were two (2) juvenile males only 2 years old:
MCM 4989, MCM 5145.
138 Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE
COOks
£6
OEE
ragiage
TEIN
OCI
6Tl
Psi
wu
Id
ie!
oll a
SS
OSP
O1S
O8P
wu
TEIN
VI
691
OT
881
LI
rol
ST
66!
PLI
PCC
col
Vsl
£0e
90T
S0T
PLI
vce
I
COU
Sid
JOJ Sv popoo oe srojotUeIed JOYjO ]][e ‘(WIDd) YSUd] Apog prepuR}s SI TS “s[vos My ULOLIPY YINOS o]eUIOJ JO syUWWOINsvoLU Apo puK s]UdUdINSeOLU [[NYS
t XIGNAddV
901
001
oll
Tu
Pld
LO
L6E
LOr
OCr
OCP
flr
Toe
ror
Ltr
OP
LP
Ll 9p
8 6b
O8b
SLb
O'6r
ob
WuuL
rel
OL
OTL
ll
601
rol
LI1
Tel
OCI
ro
OTI
OEl
Orel
OF!
OF
OT
Or!
OTL
uuu
tld
OCW
€CL
£69
OPS
WU
tld
£ Or
66E
Otr
vcr
Cot
Clr
Cc ev.
€cy
ia
Ler
LOP
Se
Sia
OP
Lb
Scr
SO
TUU
OrW
So
186
886
686
S98
OC6
96
8001
O86
SEO
LCI
0'L6
9SOL
9 COL
9'COl
0°66
TOOT
Fr to
CULL
tld
aS
Orr
Ls
WU
6c
OLE
S'8T
L6t
tou
Tid
8b
9°09
los
wu
6CcIN
PE
€CE
6 CE
wuU
Cld
L 01
£86
£0cl
wu
Sc
bcp
OP
69P
Wuuu
IId
Stel
Srl
clr
Ler
Scel
€ SEI
6 lr!
TISI
VLtl
LOSI
O'8FI
OSI
LO91
OSI
T9sI
OSS
lr
wu
STIN
HS
a ol
Vana)
tau
Ild
00
00
00
tau
Lc1
PIL
P19
HL
uu
Old
tau
LcI
COL
LSs
916
c16
bC6
VES
CES
bC6
168
0:96
iis
PLS
86
0&6
6:06
6CL
O06
O'S
wu
Old
C69 ULp iSlP
LSP 9'6E
OLS Os
LULL ragiaue COLL
9T1 scl eal
TOL ale
PLL POT
VIs SPT
Luu UU LULU
6d 8d 4d
els ceo oss
808 TS$9 (YAS)
= T99 0C9
5 (ie) LOS
= 68S IS0}S
i. 809 CLS
= Lo CHS
< 9°69 VIZ
eTs8 LL9 £19
L16 See)
96 9°69 Seo
Cc 16 802 Leo
(ara) F69 (Oa)
O16 £89 Se)
cC6 9°99 0109
Le6 199 119
6S8 oP9 PLS
UU LOU LULL
9C71 STI en |
€C8 Lo OT
PPs €6¢ 8°CT
Lige O'Ch bre
862 Cer S1e
els 5 =
68 z Cre
Pol S 1c
€C8 61k
© 6L 8'8T
8bs VIE
ses Lb OTE
818 Lor ele
cs SOF ole
88 FOr S8T
818 6 Cr 9'6T
O6L Shr 0°6C
£96 tac O'8T
008 OOF £8
TU UU UO
6d 8d 4a
$09
OPS
sro
wu
tCd
799
9°99
€CL
6r9
T89
£9
SOL
60L
£99
EL
Ded
689
eel
P89
O10L
£99
LEL
689
Cu
td
lor
£6r
Dads
Los
O'S
COP
OLS
£0S
S09
cos
in
(sa)
ay
£cs
PCS
CULL
9d
882
£99
£68
WOU
cTd
ime
SHC
C8T
fcuu
cd
9 TOL
TOOL
O'COL
£901
9°86
£46
F901
TOL
eel
cell
Til
cCll
Sstl
LOtl
aca
cell
TSOl
LO
UU
ccd
Op
CULL
sd
esl
}C UL
OCS
TOU
lcd
691
Vs
8'0T
wut
rd
v8
816
bo
bro
S68
+06
6'C6
£86
6'C6
£56
986
L96
0'COl
£66
Ono)
9'L6
£6
£6
CULT
\1cd
WANG
Lic
Sige
SPC
(oh ares
OSC
Cre
69T
OIC
99T
ESS
See
8 Pe
S9T
DEC
COU
ra
UL UT oe pue | s[Quy
cstt
SLOl
PST
Tu
Otd
oly
AUS
PSs
WOU
ed
69E1
6 9FI
Oia
O'crl
FOr
eer
8h
Lori
0'0F1
O'eS1
O6FI
Orr
09ST
OLtI
0'9OrI
Orr
OPS
00FI
TOU
Otd
PIL
SPL
LOL
TIL
Pe)
£69
99
Sel
be TL
LSL
OSL
SOL
OL
669
OL
PIL
PSL
StL
ULL
ed
SF
~
we)
isis
£01
tau
oa)
2
é
\6ld
LYrl
Ocal
OFsl
wu
cd
1601
Poll
OFT
SLIT
TOIT
Ell
Poll
1cel
0611
O9TI
olel
8 CCl
8 CEel
Ditsral|
iLL
OSI
PSII
Vsol
UU
6ld
C181
16Z1
OSS
66L1
Lgl
£821
LIs8t
TL81
TI8I
S981
L061
P98
L061
6 Esl
£rs1
Ors
L061
€ CLI
COU
cd
OTT
o80l
cctl
wut
8Id
Lego
Lcsl
O81
tau
Kal
Ls8vrl
OLS
Ses
6LSI
PS8bl
OST
SSSI
£S9]
OST
O'99T
OE9T
O91
O'OLI
OP9T
OTOL
Ol
O'E9T
OSI
uuu
8Id
8'00C
TOIT
elle
C60E
© 10
S10¢
PLO
OLIT
£80C
OCC
£9IC
DET
OSC
CRIT
OSIC
NCIC
8°CCC
SPO
Wu
d
L0¢
L0t
VIC
wu
\LId
0'06
(Ors
OOLT
Lua)
Tas
6 CC
LYE
SLE
0'6C
OFT
£6T
OLT
16c
OVE
VVC
LLT
S9C
OLE
SSTe
|©9C
O9E
POE
6c
TU
Lid
OSel
O'SPl
Och
Ser
OSEl
O'SPI1
OST
OSI
OSI
0091
Oia all
wo
Tas
esl
SLI
OSI
wuuu
Old
ool
90T
raya
OCT
0'0T
WEE
PIT
Te
ool
b CC
8'tT
10c
LOT
90T
Ol
lee
TIT
tou
od
boor WOW
066h WOW
cO8T WOW
Joquinu
uUInasnyAy
F66r WOIN
066h WOW
cO81 WOW
Joquinu
wunosny
s}nprqns/sopuoane
pSIs WOW
8L£sh WOW
borh WOW
9ssl WOW
css] WOW
Iss] WOW
9orhl WOW
101 WOW
P80e Weld
LO8T Weld
Sell Wad
9e11 Wad
peel] Wad
FIO Wad
LS6 Wid
Ico Wad
6c6 Wid
816 Wid
Jo quinu
WnasnyAy
PSIS WOW
8ZSP WOW
broth WOW
9SSl WOW
css1 WOW
Iss] WOW
96bhl WOW
101 WOW
b80¢ Wad
1981 Weld
sell Wad
Sell Wad
tell Wad
POL Wald
£86 Wid
1t6 Wid
6c6 Wad
816 Wad
qoquinu
UNO snl
S}NpV
139
Proc. Linn. Soc. N.S.W., 131, 2010
SEXUAL DIMORPHISM IN ARCTOCEPHALUS PUSILLUS PUSILLUS
APPENDIX 3.
Numbers of Individuals, Means, Standard deviations, Standard Errors and ranges of Standard Body length (SBL) and Skull Measurements in Female South
Units
Count (n)
Mean
SD
SE
Maximum
Minimum
Units
Count
Mean
SD
SE
Maximum
Minimum
SBL
cm
18.00
149.14
10.55
2.49
179.00
135.00
P16
mm
17.00
21.13
1.43
0.35
24.10
19.00
African fur seals.
17.00
60.75
4.43
1.07
71.10
50.50
M32
mm
17.00,
47.32
3.51
0.85
53.00
42.70
Proc. Linn. Soc. N.S.W., 131, 2010
140
Bacular Measurements for Age Determination and Growth in
the Male South African Fur Seal, Arctocephalus pusillus pusillus
(Pinnipedia: Otariidae)
C. L. STEwArDsON!, T. PRVAN? AND R.J. RITCHIE*”
'Botany and Zoology, Australian National University, Canberra, ACT, Australia. (Present Address, Fisheries
and Marine Sciences Program Bureau of Rural Sciences, Department of Agriculture, Fisheries and Forestry,
Canberra, ACT 2601, Australia).
"Department of Statistics, Macquarie University, NSW 2109, Australia
School of Biological Sciences, The University of Sydney, NSW 2006, Australia
*Corresponding Author: Raymond J. Ritchie, School of Biological Sciences, The University of Sydney, NSW
2006, Australia, email rrit3143@usyd.edu.au.
Stewardson, C.L., Prvan, T. and Ritchie, R.J. (2010). Bacular measurements for age determination and
growth in the male South African fur seal, Arctocephalus pusillus pusillus (Pinnipedia: Otariidae).
Proceedings of the Linnean Society of New South Wales 131, 141-157.
Morphology, relative size and growth of the baculum in 103 South African fur seals, Arctocephalus
pusillus pusillus, from the Eastern Cape coast of South Africa are described. Bacular measurements (n = 8
linear variables and mass) were examined in relation to standard body length (SBL), bacular length (BL)
and chronological age (y) using linear regression. Animals ranged from < 1 month to = 12 y. Bacular shape
was most similar to Callorhinus ursinus (Northern fur seal) and Zalophus californianus (California sea
lion). For the range of ages represented in this study, the baculum continued to increase in size until at least
10 y; with growth slowing between 8-10 y, when social maturity (full reproductive capacity) is attamed.
Growth in bacular length (BL), distal height and bacular mass peaked at 8 y; middle shaft height and distal
shaft height peaked at 9 y; proximal height, proximal width, distal width and proximal shaft height peaked
at 10 y. In the largest animal (age = 12 y), maximum bacular length was 139 mm and mass 12.5 g. Relative
to SBL, bacular length (BL) increased rapidly in young animals, peaked at 9 y (6.9%), and then declined.
Bacular mass and distal height expressed greatest overall growth, followed by proximal height, proximal
shaft height and bacular length. At 9 y, mean bacular length and mass was 117 + 2.7 4 SE, n= 4) mm and
7 = 0.7 (4) g; growth rates in bacular length and mass were 311% and 7125% (relative to age zero), and
5% and 27% (between years); and bacular length (BL) was about 6.9% of SBL. For all males 212 months,
most bacular variables grew at a faster rate than SBL and BL. Exceptions included proximal width which
was isometric to SBL; distal width and distal shaft height which were isometric to bacular length; and
proximal width which was negatively allometric relative to BL. Bacular length (BL) was found to be a
useful predictor of SBL and seal age group (pup, yearling, subadult, adult), but only a ‘rough indicator’ of
absolute age.
Manuscript received 12 October 2009, accepted for publication 21 April 2010.
KEYWORDS: age classification, age determination, Arctocephalus pusillus pusillus, baculum
morphometrics, Otariidae, Pinnipeds, South African fur seal, standard body length.
INTRODUCTION
The mammalian baculum (os penis) is found
in all carnivores, except the hyena (Ewer, 1973).
This morphologically diverse bone has received
considerable scientific attention in the field of
mammalian systematics (McLaren, 1960; Sutton and
Nadler, 1974; Kim et al., 1975; Morejohn, 1975; Lee
and Schmidly, 1977; Patterson and Thaeler, 1982;
Patterson, 1983), and has been used as an index of
age, puberty and social maturity for several species
of mammals, including pinnipeds (Hamilton, 1939;
Elder, 1951; Laws, 1956; Hewer, 1964; Bester, 1990).
The function of the baculum in carnivorous mammals
remains controversial. It may lack specific function
(Burt, 1939; Mayr, 1963) or may be adaptive in various
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
interactions of males and females during copulation,
with function differing considerably between species
(Scheffer and Kenyon, 1963; Long and Frank, 1968;
Ewer, 1973; Miller, 1974; Morejohn, 1975; Patterson
and Thaeler, 1982; Eberhard, 1985, 1996; Dixson,
1995; Miller et al., 1996, 1998, 1999; Miller and
Burton 2001). The baculum bone of carnivores is
classified as a heterotopic bone because it forms from
ossification of connective tissue (Miller, 2009). The
proximal end of the baculum is attached to the fibrous
corpora cavenosa penis.
Within the Otariidae, information on the
morphology of the baculum is available for
Arctocephalus pusillus pusillus (South African fur
seal), Arctocephalus pusillus doriferus (Australian
fur seal); Arctocephalus gazella (Antarctic fur seal);
Arctocephalus tropicalis (Sub Antarctic fur seal);
Callorhinus ursinus (Northern fur seal); Eumetopias
Jubatus (Stellers sea lion); Neophoca cinerea,
(Australian sea lion); Otaria byroni (South American
fur seal); Phocarctos hookeri (New Zealand or
Hookers sea lion) and Zalophus californianus
(California sea lion) (Chaine, 1925; Hamilton,
1939; Rand, 1949,1956; Scheffer, 1950; Mohr,
1963; Scheffer and Kenyon, 1963; Kim et al., 1975;
Morejohn, 1975; Bester, 1990; Laws and Sinha,
1993). Of these, the northern fur seal has been studied
in most detail (Scheffer, 1950; Scheffer and Kenyon,
1963; Kim et al., 1975; Morejohn, 1975).
Information on bacular growth based on bulls
reliably aged from tooth structure, or on bulls of
known age (1.e. bulls tagged or branded as pups), is
only available for Callorhinus ursinus (northern fur
seal) (Scheffer, 1950), Arctocephalus tropicalis (Sub
Antarctic fur seal) (Bester, 1990) and Arctocephalus
pusillus pusillus, South African fur seal (Oosthuizen
and Miller, 2000). A large data set of reliably aged
material is also available on the baculum of the phocid
harp seal (Pagophilus greonlandicus) (Miller et al.,
1998; 1999; Miller and Burton 2001). These studies
indicate that: (1) the baculum increases in length and
mass with increasing age; (11) bacular growth may be
fairly constant, as in the northern fur seal, harp seal
and subantarctic fur seal, or there may be an increase
in the rate of growth at puberty, as has been suggested
in the South African fur seal; (iii) there may be a
sudden increase in the rate of bacular growth when
individuals attain social maturity (full reproductive
capacity); and (iv) there is a decline in the rate of
bacular growth in socially mature bulls.
Seal baculum and testicles are used in oriental
aphrodisiac medicine and gastronomy and so there
is a legal and illicit trade in seal genitalia (Miller,
2009). Demand outstrips supply and the origin of
142
material sold is often in doubt. Bacula from South
African fur seals are part of the legal trade in seal
body parts. Other southern fur seals are not legally
hunted for body parts. It would be naive to imagine
that there is not some illicit trade in body parts from
other southern hemisphere seals and sea lions. The
other major legal source of seal body parts is from
the Harp seal (Pagophilis greonlandicus) where
illustrations, information on morphometrics, growth
and development of the baculum are available (Miller
and Burton, 2001; Muller 2009). Museums and
zoologists can be asked to identify seal body parts
by customs authorities to determine whether they are
from legally hunted species or not: morphometric
knowledge of the seal baculum is important for
conservation reasons.
Here we examine the bacula of 103 male South
African fur seals from the Eastern Cape coast of
South Africa. We provide illustrations of bacula from
the species to aid in identification. Specific objectives
were to: (i) describe the general morphology of the
baculum; (ii) quantify growth of bacular measurements
(n = 8 linear variables and mass) relative to standard
body length (SBL) (n= 89 bulls), bacular length (BL)
(n = 103 bulls), and chronological age (n = 50 bulls);
(iii) determine if the baculum is a useful indicator of
social maturity; and (iv) determine if bacular length
(BL) is a useful indicator of age and/or standard body
length (SBL). Currently there are only two reliable
means of determining the age of South African fur
seals (Stewardson, 2001; Stewardson et al., 2008).
The first is based on tagging as pups, the other is
based on dentition but the dentition method is only
valid for bulls less than about 12 y. Unfortunately, age
assignment based upon skull suture closure criteria
are known to be inaccurate and of value only for
seals > 12y in South African fur seals (Stewardson,
2001) which invalidates some early work on baculum
statistics vs. age (Rand, 1956; Mohr, 1963).
MATERIALS AND METHODS
Collection of specimens
South African fur seals were collected along
the Eastern Cape coast of South Africa between
Plettenberg Bay (34 03’S, 23° 24’E) and East London
(33° 03’S, 27 54’E), from August 1978 to December
1995, and accessioned at the Port Elizabeth Museum
(PEM), Port Elizabeth, South Africa. One animal
(PEM2238) was collected NE of the study area,
at Durban. From this collection, bacula from 103
males were selected for examination. The list of
specimens used in the present study, along with their
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
al
os
casas
Proximal Height
Width
Proximal End
Shatt Height
Length
Width
Distal End
Fig. 1 Diagram of a South African fur seal baculum indicating the variables measured (Var 1-8):
Bacular length (Var 1 or BL); Proximal height (Var 2); Proximal width (Var 3); Distal height (Var 4);
Distal width (Var 5); Three cross sectional parameters of the shaft: (1) Proximal shaft height (Var 6);
(2) Middle shaft height (Var 7) and (3) distal shaft height (Var 8). Specimen provided by P Shaugh-
nessy.
museum ascension numbers and location and dates
of collection, are listed in Stewardson et al. (2008).
Apart from specimens collected before May 1992
(n = 29), all specimens were collected by the first
author and were found dead, dying or had drowned
in fishnets.
Preparation and measurement of bacula
Bacula were defleshed and macerated in water
for 1-2 months. Water was changed regularly. Bacula
were then washed in mild detergent and air dried
at room temperature. Dry specimens were weighed
using an electronic balance and measurements (n =
8 linear variables) were taken using a vernier calliper
(to 0.1 g and 0.1 mm) following Morejohn (1975)
(Fig. 1). All bacular measurements were recorded by
the first author.
Proc. Linn. Soc. N.S.W., 131, 2010
Age determination
Of the 103 bulls in the study: (4) 40 were aged
from counts of incremental lines observed in the
dentine of upper canines (growth layer groups, GLG)
as described in Stewardson et al. (2008). Dentition-
based ages fell into 3 categories: (1) age range 1-10
y; (i1) 10 were identified as adults > 12 y (1.e., pulp
cavity of the upper canine was closed); and (iti) 53
for a variety of reasons could not be aged. None were
tagged individuals. South African fur seals older than
12 y cannot be aged from counts of growth layer
groups (GLG) in the dentine of upper canines because
the pulp cavity closes (Stewardson et al., 2008).
In studies of South A frican fur seals, 1S‘ November
is taken as the birthdate of all seals based upon
estimates of the average birthdate of pups in breeding
143
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
Table 1. The age distribution of male South African fur
seals used in the present study. Estimated age from counts
of incremental lines observed in the dentine of upper ca-
nine (n= 40). An additional 10 males were > 12 y, i.e., pulp
cavity closed. Pups were greater than one month of age.
measurement error. The Wilcoxon sign-rank
test was used on the differences to test H,:
median = 0, versus H,: median # 0.
Bacular length (BL) expressed in relation to
standard body length (SBL)
Standard body length (SBL) is defined as
colonies (Rand, 1949; Oosthuizen and Miller, 2000).
For this study, the following age groups were used:
pup (< 1 months to 6 months); yearling (7 months to |
y 6 months); subadult (1 y 7 months to 7 y 6 months);
and adult (> 7 y 7 months) (rounded to whole years
in Table 1) (see Stewardson et al., 2008, 2009). No
individuals of 2 y to 3 y were available. Data on
very old bulls that had been tagged as pups were
not available. The estimated longevity of bull South
African fur seals is about 20 y based primarily on
zoo animals (Wickens, 1993). Currently, examination
of tooth structure is the most precise method of age
determination in untagged pinnipeds; however,
counts are not without error. For information of the
reliability of this method see Oosthuizen (1997) and
Stewardson et al. (2008).
The limitations of age determinations based upon
dentition become apparent if one realises that it would
be reasonable to assume that the longevity of South
African fur seal bulls in the wild would be at least
15 y (based upon documentation on the Australian
fur seal, A. pusillus doriferus; Arnould and Warneke,
2002), which implies that dentition can only age male
South African fur seals up to only about 2/3 of their
total potential lifespan.
Statistical analysis
Bacular measurement error
Duplicate measurements of bacular length were
taken from 50 randomly selected bacula to assess
144
the length from the nose to the tail in a straight
line with the animal on its back (Committee
on Marine mammals, 1967). Growth in BL,
relative to standard body length (SBL), was
calculated as follows, using paired samples
only:
BL (mm) /SBL (mm) x 100%
As the approximate variance of the ratio
estimate is difficult to calculate, percentages
must be interpreted with caution (Cochran,
1977, p. 153).
Bacular growth relative to age zero, RGR
ne
Percent change in bacular measurement
at age t, relative to value at age zero, was
calculated as follows:
[(Y.-Y,)/Y,] x 100%
where, Y, = mean bacular measurement from pups <
1 months of age (age zero), and Y, = mean bacular
measurement for age t (age class in y).
Bacular growth relative to the previous year (annual
bacular growth), RGR Y_ ,
The percent change in value at age t, relative to
the value at age t-1, was calculated as follows:
((Y-Y,,)/Y,,] x 100%
where, Y, = mean bacular measurement for age (t),
and Y,, = mean bacular measurement for age t-1
(between years). RGRs were calculated for bulls that
were 7-10 y.
Bacular length (BL) as an indicator of SBL and age
The degree of linear relationship between Log,
(BL), Log, (SBL) and Age (y) was calculated using
the Spearman rank-order correlation coefficient.
Linear discriminant function analysis (Mahalanobis
squared distance) was used to predict the likelihood
that an individual seal will belong to a particular
age group (pup, yearling, subadult, adult) using one
independent variable, bacular length (see Stewardson
et al., 2008, 2009 for further details).
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
Bivariate allometric regression
The relationship between each bacular
measurement (Var 1 to 9) and: (1) SBL, (i) BL,
and (iil) age (y), was investigated using linear
regression, semi-log plots (Log, y = mx + b) or the
log/log logarithmic transformation of the allometric
equation, y = ax’, which may equivalently be written
as Log. y = Log, a + b.Log, x. For most analyses the
three one month-old pups were not included (hence
n = 37). ‘Robust’ regression (Huber M-Regression)
was used to fit straight lines to the untransformed or
transformed data. The degree of linear relationship
between the transformed variables was calculated
using the Spearman rank-order correlation coefficient,
r (Gibbons and Chakraborti, 1992). Testing of model
assumptions, and hypotheses about the slope of the
line, followed methods described by Stewardson et
al. (2008).
Statistical analysis and graphics were
implemented in Minitab (Minitab Inc., State College,
1999, 12.23); Microsoft ® Excel 97 (Microsoft
Corp., Seattle, 1997) and SPLUS 7.0 (MathSoft, Inc.,
Seattle, 2005, version 7.0).
RESULTS
Bacular measurement error
Of the 50 bacula that were measured twice,
measurements were reproducible at the 5% signifi-
cance level (p-value = 0.052).
Bacular morphology
Bacular length (BL) and mass ranged from 26.6
to 139.3 mm and 0.1 to 12.5 g, respectively (Table
2),
The youngest animals in the sample were < 1
month of age. In these individuals, the baculum was
short, thin and rod-like, with no obvious distinction
between the proximal and distal ends (Fig. 2a and 2b).
The shaft was slightly curved anteriorly (variable).
In yearlings, the baculum increased substantially
in length and mass (Table 3). The distal end was
slightly rounded but, there was no sign of bifurcation
(Fig. 2c).
In subadults, most bacula curved upwards at the
distal end (i.e., superiorly, see Fig. 2d). At the distal
end of the baculum, there were two narrow projections
(knobs): a well-developed ventral knob and a less
prominent dorsal knob (Fig. 2d). In older subadults,
the ventral knob extended upwards and outwards
forming a double knob (variable). The proximal end
of the bacula was bulbous in all bulls > 4 y.
Proc. Linn. Soc. N.S.W., 131, 2010
In adults (> 8 to 9 y) the baculum was well
developed, with pronounced thickening of the
proximal end. Contrast Fig. 2d which is a 7 year old
subadult with Fig. 2e which is a 10 year old (Fig. 2).
At the bifurcated distal end, the ventral knob usually
extended further than the dorsal knob. In older males,
the baculum was more robust, but not necessarily
longer. Small osseous growths were commonly found
on the proximal end of the baculum (n = 18 subadult
and adult bacula) creating a rough surface where the
fibrous tissue of the corpus cavernosum penis attached.
In some older specimens (n = 16 bacula), small knob-
like growths (usually 1 or 2) were observed along the
edge of the urethral groove, at the proximal ventral
surface of the baculum.
Bacular length expressed in relation to SBL
Relative to SBL, BL increased rapidly in young
animals, peaks at about 9 y (6.9%), and then declines
in old bulls = 12 y, i.e., adults 8 to 10 y, mean 6.6
+ 0.122% (n = 13) vs. adults > 12y, 6.09 + 0.32%
(n = 9); t-test p< 0.01. More detailed relative growth
patterns for subadults, adults and old bulls could
not be established because the sample size is too
small and SBL was not available for all specimens
(SBLs for 12 animals drowned in fishnets were not
recorded because rough conditions at sea precluded
measurement of SBL).
Bacular growth relative to age zero, RGR Y,
Percent change in value of bacular measurement
at age ¢, relative to value at age zero, is presented
in Table 4. In yearlings, bacular mass was the most
rapidly growing variable, followed by bacular length,
proximal height, distal height, proximal shaft height,
proximal width and distal shaft height/middle shaft
height. Distal width showed little sign of growth.
Growth of bacular variables continued to increase
until at least 10 y, with bacular mass, middle shaft
height and distal shaft height expressing continued
growth in bulls => 12 y. Bacular mass and distal
height expressed greatest overall growth, followed by
proximal height, proximal shaft height and bacular
length (Table 4).
Bacular growth relative to the previous year,
RGRY _,
Percent change in value of bacular measurement
at age ¢, relative to value at age 7-1, for bulls 7-10 y,
is presented in Table 4. Percent increment in bacular
length, distal height and bacular mass peaked at 8 y;
middle shaft height and distal shaft height peaked at
9 y; proximal height, proximal width distal width and
proximal shaft height peaked at 10y.
145
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
Age
group
)
' Pisyey 22 || (yee beope sane el Theanine eer se | il O)ee |) iT Se
1.6 0.5 0.3 0.3 0.2 0.2 0.2 0.1 0.0
(9.6)- (31.5) | (12.5) | (24.7) | (18.3) | (13.6) | (15.7) | (7.9) | (0)
; 9.0% 12.3% | 7.8% | 5.9% | 8.3% | 7.7% | 6.8% | 0.4%
Age Var 1 : Var Var Var Var Var
<
AGL as || Jses. | AL Aes |) 2G)ee P= || 3:0 | 2 Sie Set Ors
1.7 0.1 0.1 0.2 0.04 0.1 0.1 0.2 0.03
Yearling
Subadult
Adult
n
3
5
(8.0) - (7.7) (6.6) (15.8) | (5.9) (5.0) (12.2) | (18.2) | (23.6)
: 7.3% 8.8% 6.1% 3.6% | 6.2% | 5.2% | 4.6% | 0.6%
4 SIS 06:6) Shea aks
3 971+ | 94+ 7.7+ 94+ AV ae || Fes || S32 | Si)2= |) aiabee
4.6 2.5 0.9 0.6 0.8 0.6 0.2 0.2 0.4
(8.2)- (45.3) | (20.9) | (10.5) | (31.0) | (13.6) | (4.6) (8.4) (21.2)
: 9.7% 7.9% 9.7% 4.3% | 7.2% | 6.0% | 5.1% | 3.5%
3 : : 10.9+]/3.94 | 71+ |54+ | 45+ |] 3.14
0.1 0.6 0.9 0.2 0.1 0.1
(0.7) (20.2) | (17.9) | (5.2) (3.1) (2.3)
10.9% | 3.9% | 7.1% | 5.4% | 4.5% | 3.1%
7 10.74} 40+ | 72+ | 63+ |53+4 | 41+
0.6 0.2 0.3 0.3 0.2 0.4
(17.8) | (17.5) | (14.8) | (13.3) | (14.3) | (34.0)
10.5% | 4.0% | 7.1% | 6.2% | 5.3% | 4.0%
4-7 17 9.3 + T5+ 10.3+ | 4.0+ | 7.14 | 614 | 5.1+ | 3.74
: 0.8 0.3 0.4 0.2 0.2 0.2 0.2 0.3
(8.7) - (34.6) | (17.5) | (17.5) | (20.5) | (14.4) | (12.5) | (13.9) | (33.1)
; 9.3% 7.5% 10.3% | 4.0% | 7.1% | 6.1% | 5.1% | 3.7%
111.4 113= | 914= IWWrze | 4.332 | SOz | 6M | S62 |) S7/se
+ 3.] 0.8 0.6 0.5 0.1 0.3 0.2 0.2 0.5
(7.8) - (19.0) | (18.5) | (12.3) | (9.5) (11.1) | (8.7) (8.4) (23.9)
; 10.8% | 8.4% 11.0% | 3.9% | 7.2% | 6.1% | 5.0% | 5.1%
n 116.9 10.4+ | 10.8+ | 12.44 | 49+ | 814 | 76+ | 63+ | 722
+27 1.8 1.6 0.9 0.7 0.5 0.3 0.2 0.7
(4.6) - (35.5) | (29.3) | (14.5) | (29.2) | (12.8) | (7.9) (7.8) (18.4)
; 8.9% 9.2% 10.6% | 4.2% | 7.0% | 6.5% | 5.4% | 6.2%
3 117.8 14.0+ | 13.54 | 13.2+ | 61+ | 10.6 Qise | GSse || WSs=
+2.9 0.8 1.9 0.5 0.4 +03 | 04 0.2 0.6
(4.3) - (9.7) (24.5) | (6.2) (12.5) | (4.8) (8.1) (4.7) (14.1)
y 11.9% | 11.4% | 11.2% | 5.2% | 9.0% | 6.9% | 5.5% | 6.5%
15 114.2 11.6+ | 10.6+ | 125+] 48+ | 86+ | 73+ | 60+ | 65+
+2.0 0.7 0.7 0.4 0.3 0.3 0.2 0.1 0.4
(6.6) - (23.1) | (26.4) | (11.5) | (22.0) | (15.4) | (10.6) | (9.6) (23.2)
: 10.2% | 9.3% 10.9% | 4.2% | 7.5% | 6.4% | 5.2% | 6.7%
SH lls AT SIstgh ol Aaa cs eg Seal: ae $16 talla6 Ges NSE
7 +3.8 0.8 0.7 0.7 0.5 [8], 0.6 0.3 0.9
(10.7)- (22.6) | (20.9) | (17.3) | (28.4) | (17.2) | (23.6) | (12.5) | (34.2)
; 10.1% | 8.9% 11.7% | 4.5% | 8.8% | 7.6% | 5.8% | 7.3%
Totalorei/et ili ES01> ile SON aa ASD |"50. MRSS OMe EC Ei es
Mean for males > 200 cm WI 13.14} 9.94 14.4+ | 5.0+ 10.5 oes |l 7 jlse 10.9
(n= 7) +28 0.3 1.0 0.4 0.3 +05 | 0.3 0.3 + 0.5
[Maximum value in brackets] [139.3] | [14.0] | [13.7] | [15.7]
146 Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
Fig. 2 Size and shape of the South African fur seal baculum in relation to age group:
a. pup (PEM2020, 26.6 mm); b. pup (PEM2024, 31.6 mm); c. yearling (PEM2191, 50.7 mm); d. sub-
adult, 7-y-old (PEM2053, 93.3 mm) and e. adult, 10-y-old (PEM2087, 123.3 mm).
Bacular length as an indicator of age
The plot of Bacular length (BL) vs. Age (y) is
shown in Fig. 3. For animals 1-10 y, bacular length
was highly, positively correlated with age (y) (r
= 0.825, n = 38; Fig. 3). However, after fitting the
straight line model, the plot of the residuals versus
fitted values was examined, and the straight line
model was found to be inadequate (the residuals were
not scattered randomly about zero, see Weisberg,
1985, p. 23). Thus, strictly speaking bacular length
could not be used as a reliable indicator of absolute
age based on a simple linear model but could be used
as a rough indicator of age.
For the range of ages available in this study (Table
2), the coefficient of variation in bacular length for
young males 1-5 y (36.8%) was considerably higher
than in older males (8-10 y, 6.6%; > 12 y, 10.7%).
Although bacular length was not a good
indicator of absolute age, it was more accurately a
‘rough indicator’ of age group. When bacular length
is known, the following linear discriminant functions
can be used to categorise each observation into one of
Table 2 (LEFT). Summary statistics for bacular variables (1 - 9), according to age (y) and age group.
Data presented as the mean + SE, followed by coefficient of variation in round brackets, and bacular
variable expressed as a percentage of bacular length. Maximum value of each variable (males of un-
known-age) is also presented. All measurements are in mm, apart from bacular mass (g).
Variables: 1. Bacular length (BL); 2. Proximal height; 3. Proximal width; 4. Distal height; 5. Dis-
tal width; 6. Proximal shaft height; 7. Middle shaft height; 8. Distal shaft height; 9. Bacular mass.
Number (n) is the number of bacula from individuals where their age had been determined based on
dentition. Sample size given in square brackets where this does not equal total sample size. Mean value
of variable + SE for the 7 largest males (© 200 cm, SBL) of unknown-age; maximum value in brackets.
Proc. Linn. Soc. N.S.W., 131, 2010 147
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
Table 3. Growth in mean bacular length (BL) relative to mean standard body length (SBL). Number (n)
shows the number of canine aged animals where both BL and SBL were recorded. Of the 50 canine aged
animals, SBL was not recorded for 12 animals, i.e. n= 38. Sample size is given in square brackets where
this does not equal total sample size. Bacular length (BL) values are mean + SE in mm. SBLis expressed
as mean + SE in cm. Relative bacular length (RBL) is defined as 100% x BL (mm)/SBL (mm).
fener bacular
TEARS A0)) : ee length (BL) (mm)
mw a hee
Yearling 47.8 + 1.7 [5]
elative Bacular
Mean SBL (cm)
ene ey
R B } ( A B U
aeeEEenate isos se
69.0 + 2.5 [3] 4.1% [3]
90.6 + 2.7 [5] 5.3% [5]
Subadult 86.6
Adult 8
Total
four age groups (pups, yearlings, subadult, adults):
Pup = -5.50 + 0.39 x BL
Yearling = -15.53 + 0.65 x BL
Subadult = -67.25 + 1.35 x BL
Adult = -87.77 + 1.54 x BL
where, BL = bacular length (mm); Age Classes: pup,
yearling, subadult and adult. The seal is classified into
the age group associated with the linear discriminant
function which results in the minimum value (see
Stewardson et al., 2008, 2009). Of the 50 animals in
this study, 86% were correctly classified using this
method (Table 5).
Bacular length as an indicator of SBL
The plot of Log, (BL) vs. Log, (SBL) is shown in
Fig. 4. Log, Bacular length (BL) was highly positively
148
102.2 (1 measured)
a CN CCT CE
7m _fessessir [sree [erm
fe [rmossamn [amar [oon
CC
38 38
145.0 (1 measured)
6.9% [3]
6.3% [3]
6.6% [13]
6.1% [9]
linearly correlated with SBL (r = 0.877, n = 86; Fig.
4) on a plot of SBL (cm) vs. BL (mm) using robust
Huber M Regression. When bacular length is known,
the following equation (linear least squares fit; Log.
transformed data) can be used to predict Log. (BL);
Log, (BL) = -2.062 (+ 0.247) + (1.3142 + 0.0493) x
Log, (SBL)
where, the Spearman rank-order correlation was
0.877. M-estimate was not significant for bias (p =
0.0945) but LS-estimates for bias were significant (p
= 0.00048).
Bivariate allometric regression
Spearman rank-order correlations show that
bacular variables were significantly (p < 0.01) with
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
Table 4. Growth in bacular variables (1-9) relative to the mean value of bacular measurement (i) at age
zero, RGR YO and (ii) from the previous year, RGR Yt-1. Growth in SBL is also given. All measure-
ments are in mm, apart from the SBL (cm) and the bacular mass (g).
Variables: 1. Bacular length (BL), 2. Proximal height, 3. Proximal width, 4. Distal height, 5. Distal
width, 6. Proximal shaft height, 7. Middle shaft height, 8. Distal shaft height, 9. Bacular mass. n is the
number of canine-aged animals. SBLs of 12 animals were not recorded. Values for growth relative to age
zero are presented on the left side of the relevant columns, i.e. [(Yt-Y0)/Y0] x 100 where Yt is the mean
value at time t and Y0 is the value at time zero. Values for growth relative to the previous year are pre-
sented on the right hand side of the relevant columns. For animals 7 to 10 y of age, i.e. [(Yt-Yt-1)/Yt-1] x
100 where Yt-1 is the mean value for the previous year class and Yt is the mean value at time t. Sample
sizes are given in brackets where this does not equal the total sample size. Instances where growth could
not the calculated are marked (*) and there are two cases where the calculated growth is negative (adult
age 7y; Var 4 and adult age 9y; Var 2).
Age Age Varl |; Var | Var | Var | Var
n SBL
Class (y) (BL) | 2 3 4 5
Subadult | 4 1 99 204 106 | 89 227 | 68 149 150 128 | 2300
5 3 *(0] 241 266 20a S22 152 196 | 164 157 | 3300
6 218 | 91 386 131 | 200 145 133 | 2950
7 Mise i iliss | SAMs | 14's 176; | 3964;
QQA | WB | slp Sail 18.8 | 33.2
341; | 169; 5
ous ISS) 2H!
9 304; | 209;
-8.3 | 14.9
447; | 285;
tl 35.3 |, 2A
S12 Oj les yy) 2o7 343 189
each other (Table 6). Distal width (Var 5) with
proximal width (Var 3) had the lowest correlation (r
= 0.67) but most equal or exceed r = 0.80. Plots of all
the data used for the bivariate allometric regressions
can be found in Stewardson (2001). In the present
study, the slope and intercept values and correlation
coefficients (1) are shown in Tables 7, 8 and 9.
Regression of bacular measurement on SBL
Of the 103 seals in the study, 86 were used in
regression analysis for the natural log of baculum
measurement on Log, (SBL). All pups (n = 3) were
Proc. Linn. Soc. N.S.W., 131, 2010
excluded from the regression analysis, and SBLs for
12 animals had not been recorded (see above).
There was little difference between the ordinary
least square straight lines fitted to the data, and the
‘robust’ least squares straight lines fitted to the same
data. The ‘robust’ straight line equations for regressing
log of baculum measurement on log of seal length are
given in Table 7. All bacular variables were highly,
positively correlated with SBL, r => 0.68. Relative to
SBL, growth in distal height, distal width, proximal
shaft height, distal shaft height and bacular mass
was positively allometric; and proximal width was
149
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
140
120
100
=” a0
=
ra
60
40
6 8 10
Age (y)
Fig. 3 Bivariate plot of Baculum Length (BL) (mm) vs. age (y) using Robust MM Linear regression.
The fitted line was BL = 48.63 (+ 10.39) + (7.678 1.346) x Age with a Spearman rank-order correlation
of 0.825. The M-estimate and LS-estimate for bias were not significant. Robust MM Linear regression
could also be run to predict Age (y) from BL. The fitted line was Age = -4.016 (+ 1.166) + 0.108 (+ 0.0111)
x BL.
isometric (Table 7). Regression slopes for bacular
length, proximal height and middle shaft height all
had significant positive slopes > 1 (Table 7).
Value of bacular measurement on bacular length
Of the 103 seals in the study, 100 were used
in regression analysis for natural log of baculum
measurement on bacular length. All pups (n = 3) were
excluded from the regression analysis.
All bacular variables were highly, positively cor-
related with bacular length, r > 0.7 (Table 8). Relative
to bacular length, growth in distal height, proximal
shaft height and proximal height was positively
allometric relative to bacular length; distal width and
distal shaft height was isometric; and proximal width
was negatively allometric (Table 8). Regression slopes
for middle shaft height and bacular mass scaled with
positive slope (Table 8). The slope for bacular mass
was considerably steeper than for other variables.
150
Value of bacular measurement on age
Of the 40 seals aged from upper canines, 37 were
used in regression analysis for the natural log of a
baculum measurement versus age. As above, all pups
(n = 3) were excluded from the regression analysis.
Overall, the plots of log bacular measurements
versus log SBL were better described by linear
relationships than the plots of log, bacular measure-
ments versus age (see Griffiths et al., 1998, p. 126).
Fig. 3 shows a plot of BL vs. Age (y); data for this
and other fits are shown in Table 9. Proximal height
vs. Log, (SBL) was the only variable that roughly
resembled a straight line.
DISCUSSION
Bacular size
In South A frican fur seals (Arctocephalus pusillus
pusillus) from the Eastern Cape coast, maximum
Proc. Linn. Soc, NSW. Asi, Zone
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
Table 5. Discriminant analysis for male seal age group (pup, yearling, subadult and adult) inferred from
bacular length. Number (n) is the number of animals aged from counts of incremental lines observed in
the dentine of upper canines, n= 50. Percentage of animals correctly classified into age group is given in
brackets. Animal classified as adults includes animals > 12 y.
Known Age Group Classification into age group
(18 month <
age<7y6
month)
Yearling
(7 month
<age< 18
Pup
Yearling
Subadult
Adult
Ln (BL, mim)
4.3 4.5 47 4.9 5.1 5.8 5.5
Ln (SBL, cm)
Fig. 4 Bivariate plot of Loge (BL) vs. Loge (SBL) using Robust MM Linear regression. The fitted line
was Loge (BL) = -2.062 (+ 0.247) + (1.3142 + 0.0493) x Loge (SBL) with a Spearman rank-order correla-
tion of 0.877. The M-estimate was not significant for bias (p = 0.0945) but the LS-estimate for bias was
significant (p = 0.00048).
Proc. Linn. Soc. N.S.W., 131, 2010 151
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
Table 6. Spearman rank-order correlation coefficients for log bacular variables. Variables: 1. bacular
length (BL); 2. Proximal height; 3. Proximal width; 4. Distal height; 5. Distal width; 6. Proximal shaft
height; 7. Middle shaft height; 8. Distal shaft height; 9, bacular mass. Two distal width measurements
were not recorded because specimens PEM2049 and PEM2134 were damaged hence Var 5 has only 101
records. All correlations are significant at the 1% level (2-tailed), i.e. p < 0.01.
Sm et tt er
es ei
(BL)
Var2 | 0.82 035 6[084 |080 | 0.85
Var3 | 0.71 0:76, 16.750. |\lormomnalaorm
Var4 | 0.90 0.86 | 0.89
Var5 | 0.80 0.79 | 0.80
Var6 | 0.88 1.00 | 0.94
Var7 | 0.92 0.84 | 0.75 094 | 1.00 [0.96 | 097 |
Var8 | 0.90 0.80 | 0.70
Var9 | 0.95 0.85 | 0.77 0.94 | 097 | 0.95 | 1.00
bacular length we found in the present study was 139.3
mm and mass was 12.5 g; however bacula up to 141
mm (Oosthuizen and Miller, 2000) and 16.8 g (Rand,
1949) have been reported for South African fur seals
from other areas. Baculum length was similar to that of
the Northern fur seal (Callorhinus ursinus) (Scheffer,
1950) and the harp seal (Pagophilus greonlandicus)
(Miller and Burton, 2001; Miller 2009), which is a
phocid seal. As with other Otariidae, bacular length
of the South African fur seal is considerably smaller
(proportionately to standard body length, SBL) than
that of most Phocidae and the Odobenidae (Scheffer
and Kenyon, 1963; Miller and Burton, 2001).
No systematic quantitative study seems to have
been made of the growth with age of the baculum
of the Australian fur seal (Arctocephalus pusillus
doriferus) or the New Zealand fur seal (Arctocephalus
forsteri). Basic morphometric data on the bacula of
Australian and New Zealand fur seals do not appear
to be readily available (Scheffer and Kenyon, 1963).
At present it would be very easy to pass off illegally
obtained bacula from Australian and New Zealand
seals as legal South African material.
Bacular shape
Although detailed information on the morphology
of the otariid bacula is sparse, bacular shape was most
similar to the Northern fur seal and California seal
[52
lion (Kim efal., 1975; Morejohn, 1975; King, 1983).
For example, in Arctocephalus fur seal species,
Northern fur seal and California seal lion, the adult
bacular apex consists of a dorsal and a ventral knob.
When viewed anteriorly, the knobs are parallel sided
(Arctocephalus species and the California sea lion),
or resemble a figure-of-eight in the California sea
lion. Apical keels (lateral expansion of the apex) are
present on the baculum of some California sea lion
individuals, yet absent in both Arctocephalus species
and the Northern fur seal (Kim ef a/., 1975; Morejohn,
1975).
Bacular length (BL) as an indicator of Standard
Body Length (SBL) and age
As with other species of pinnipeds, there is
considerable variation in BL with age, especially in
younger animals (Rand, 1949; Scheffer, 1950; Bester,
1990; Oosthuizen and Miller, 2000).
In male South African fur seals, BL was found
to be a ‘rough indicator’ of SBL and age group, but
not of absolute age. The classification criteria for
age group, and SBL, developed in this study will be
particularly useful when teeth are not available for
age determination; a seal is decomposed/scavenged
(total SBL cannot be measured) or because the skull is
incomplete/absent (total SBL cannot be extrapolated
from skull length); or museum records have been ©
misplaced or destroyed. As more specimens become
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
90'0 + 9E'1 cree
60°0 + 8e 1 y0'0 + 79'S-
Peeee VN VN (10'0 >) 90°0 + €S'I
90°0 + 0S'I
(10°0 >) 080
(10'0 >) 78°0 60°0 + 00°72
su 8/0 8 [#u-H (10'0 >) 890 60°0 + £0'I
VN VN VN (10°0 >) 840 60:0 + 7ST
VN VN VN (100 >) 880 70'0 + €7'I
anyjea-d jp yale (sonjea-d) I (mm) Sane
Es AaQoulo[[V
890 = IS IC- uIn[NdEq JO SseIA, *6
JYSIoY WeYs [eISIG “8
870 = C6'S- JYSIOY IEYS [PPA *L
JYSIOY yeys [VUIXO1g
6c 0 ¥ 6S S-
PIM [eIsIq
9r'0 F 88°L- WYSIOY [LISIG “PF
8v'0 + CI €- U}PIM [VUNIXO1g
cv 0 = 8S S- 98 JYSsIoY [VUIXOIg
COW # LO 98 (Ta) unynoeg Jo yysuaT
as + ii
(q) 3dao.103uT
uOIssatsed IvoUrT | a[quliva Juepusdsg
(9 o[qB], 99S) SusuTIDeds 0M} WO poANsvoU 9q
JOU P[NOd s}UIWIAANSeIU Y)PIA [eISIP gsnvdEq PR = U (C IBA) YIPIA [LISIP AO] x “IAJOWLOST SI YJMOAS ILOJIAIY) SOA] IUBIIUSIS %C IY} Iv [}{ JO ANOALY
Ul ‘OFT 990f91 JoUUD 9M ‘CQ'() < SUA ONTVA-d JY} DUIS (SU) JUBIIUSIS JON] “JOU JOU 3.19 (UI) BUTT BY) JO odoys sy) ynOGe sasayjodAY 3s} 0} posmMbos suoy
-duinssv Japour ay} ssnedeq a[quarjdde jou s}s93 ‘WN ‘OUT, pe73y 24) Jo adojs pourUitsjJ9p 9Y} SI WI “(pofie}-Z) [OAT] %] IU} IV JUBIIUSIS 919M SUOILIOIIOD
ILV ‘}U919HJI09 UOLII1109 19p.10-yuLA UBULIvads si 1 ‘(e[NIvq 98 = U “‘9°l ‘pop10I0 JOU 9.10 So[VU p] WOAJ STS pue ‘sisAjvue WOA popnydxo o10M sdnd ¢
9Y}) 9sv-UMOUYUN JO s[VUIUL 10} puL S[LUTIUL pose-oUTULD WO, BNI JO AIQUINU [&}0} JY} 10}J SI (U) AZquINU 9Y[, ‘(WID) YISuUe] Apog [vos so] UO (UI) JUOT
-IANSLIU AV[NIG SO] AOJ ATVIUIOT[Y PUL SJUIIDYJIOI UOL[I110D 19p.10-yuLA UvUAvadS ‘(q + xu = A) suOLENbo sul] JYSIV.Ws SoreNbs ysvay Jsnqoy, £ 9[quy,
153
Proc. Linn. Soc. N.S.W., 131, 2010
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
Table 8. ‘Robust’ least squares straight line equations (y = mx + b), Spearman rank-order correlation coefficients and allometry for loge (bacular meas-
urement) on loge (bacular length). Number (n) is the total number of bacula for canine-aged animals and animals of unknown-age (the 3 pups were
excluded from analysis, i.e., n = 100 bacula). * For distal width (Var 5) n = 98 because two specimens were damaged (see Table 6). r is the Spearman
rank-order correlation coefficient. All correlations are significant at the 1% level (2-tailed). Test not applicable (NA) because model assumptions required
to test hypotheses about the slope of the line (m) were not met. Not significant (ns) since the p-value was > 0.05, we cannot reject H_,, in favour of H, at the
5% significance level; therefore growth is isometric.
oO
D_pendent variable Linear regression |Allometry
Intercept (b) Slope (m) Alternative fe
n r (p-values) 2 df p-value
+ SE + SE Hypothesis
2. Proximal height 1000) 9] Ssi0126" 421 o106 oso eo emt iss |e o0rs
3. Proximal width 100 | -1.52+£0.29 | 0.79 £0.06 oe GOO). iS |, 98 ons
4. Distal height iy SOO Tio 20e (wes e0o semi ss |-oo |
5. Distal width O8e -3.61 + 0.26 1.08+ 0.06 | 0.79 | (<0.01) 96 0.15 ns
. Proximal shaft height 100 -3.30 + 0.17 1.16 + 0.04 (< 0.01) < 0.01
oor [Eeooh | NA Ns
0.89 | (< 0.01) lbs saa == II 98 0.15 ns
0.94 | (<0.01) NA NA
Table 9. ‘Robust’ least squares straight line equations and Spearman rank-order correlation coefficients for log (bacular measurement) vs. age (y) and
for log (weight) vs. age (y). n is the total number of bacula for canine-aged animals (only animals 1 to 10 y were included in the analysis, hence n = 37).
SBLs for 11 aged males were not recorded. ris the Spearman rank-order coefficient. All correlations were found to be significant at the p < 0.01 level (2
tailed).
Slope (m) + SE r (p-values)
1.17 + 0.03
1.05 + 0.04
3.49 + 0.06
-3.52 0.15
-3.18 + 0.29
-14.66 + 0.29
. Distal shaft height 100
. Mass of baculum | 100
Fe
ie)
Z Z,
> >
6
7. Middle shaft height | 100 _|
8
9
Dependent variable ‘Robust’ Log-Linear regression
n Intercept (b) + SE
_ Length of baculum (BL) 37 3.88 + 0.05 0.10+0.01 0.83 (< 0.01)
. Proximal height 137 1.13 £0.08 0.15 + 0.01 0.67 (< 0.01)
Proximal width 1.31 +£0.09 0.11 +£0.01 0.78 (< 0.01)
. Distal height 1.10 + 0.10 0.17 £0.01 0.76 (< 0.01)
37)
. Proximal shaft height 1.05 + 0.06 0.13+ 0.01 0.74 (< 0.01)
37
. Middle shaft height (37 | 0.89+0.13 0.13 + 0.01 0.85 (< 0.01)
. Distal shaft height 137 | 0.82 + 0.06 0.11 +0.01 0.79 (< 0.01)
. Mass of baculum Cree Besar 0.37 + 0.02 0.87 (< 0.01)
Standard body length (SBL) 26 4.46 + 0.04 | 0.08 + 0.01 0.83 (< 0.01)
]
2
3
4
5. Distal width 0.45 + 0.07 0.13 £0.01 0.68 (< 0.01)
6
7
8
9
Proc. Linn. Soc. N.S.W., 131, 2010
154
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
available, the classification criteria would be expected
to become more precise. Statistics on age vs. bacular
length show that bacular length can be used as a
rough indicator of age (Fig. 3) and show that it is a
better indicator of age than Standard Body Length
(SBL) in terms of correlation coefficient (r) and error
of the predicted age (Stewardson et al., 2009). More
determinations of bacular length in tagged bulls of
known age could prove it to be a very useful method.
Bacular growth
In male South African fur seals, growth of the
baculum is a differential process with most variables
growing rapidly relative to SBL and bacular length
(BL). Two variables were isometric and one was
negatively allometric, relative to bacular length,
indicating that the adult baculum was not simply an
enlarged version of the juvenile baculum (see Fig.
Dy,
Growth changes in BL and mass described in
this study generally support findings reported by
Oosthuizen and Miller (2000) and are also similar to
those reported for the harp seal (Miller and Burton,
2001) which is a phocid seal. In this study, based
primarily on animals collected from the south and
south-west coast of southern Africa, growth in BL
took place rapidly up until 5 y; peaked at 9-10 y; and
then slowed. Our findings could not be compared
to those of Rand (1956) because, in the latter, age
was estimated from cranial suture closure which has
subsequently been shown to be an unreliable indicator
of absolute age in this species, particularly for animals
> 12 y (Stewardson et al., 2008).
The biological significance of bacular growth
patterns
In male South African fur seals, a growth spurt in
BL occurs at 2-3 y (Rand, 1949; Oosthuizen and Miller,
2000), when males attain puberty (Stewardson et al.,
1998). Unfortunately, we have very scanty details on
the life history of South African fur seals during the
dispersive juvenile stages of their life. After puberty,
the baculum continues to increase in length with
increasing age, approximating full length at about 9 y
(Oosthuizen and Miller, 2000; present study). Bacular
dimensions, other than length, approximate full size
between 8-10 y (present study), when most males have
attained full reproductive capacity (Stewardson et al.,
1998). Although males can sire offspring at a young
age (e.g., at 4 y in captivity; Linda Clokie-Van Zyl,
pers. comm.), bacular growth is geared to coincide
with the attainment of social maturity, presumably to
enhance the effectiveness of copulation.
Proc. Linn, Soc, N.S2W., 13:1, 2010
Socially mature male South African fur seals:
(1) may achieve a high level of polygyny at large
colonies (David, 1987); (11) usually copulate once
with each harem female, 5-7 days postpartum during
a brief breeding season (November to late December)
(David and Rand, 1986); and (iii) usually exhibit
brief intromission duration (Stewardson, pers. obs.).
In such males, the baculum is therefore large enough
to provide sufficient mechanical support for insertion
and repeated copulations (with potentially numerous
females within a short period of time), and may assist
in deeper penetration. The ornate apex presumably
serves to stimulate the vagina of the female (Eberhard,
1985, 1996). However, the function of the apex in this
Species remains unclear considering that: (1) female
South African fur seals are not “induced ovulaters’
like cats; (ii) copulation occurs when the female is
sexually receptive and (111) sperm competition is
weak (Stewardson et al., 1998).
CONCLUSION
Data presented in this study provide more detailed
information on the morphology of the South African
fur seal bacula than earlier descriptions given by
Rand (1956) and Mohr (1963), based on smaller data
sets and more dubious age estimates. Oosthuizen and
Miller (2000) used a larger data set than the present
study but did not attempt a detailed analysis of bacular
morphometrics. Our study provides new information
on the patterns of bacular growth in relation to age and
SBL (Oosthuizen and Miller, 2000), and demonstrate
that bacular length is a ‘rough indicator’ of SBL and
age group. Similar overall conclusions have been
drawn from analysis of larger data sets available for
the harp seal (Miller et al., 1998, 1999; Miller and
Burton, 2001) which is a member of the phocidae
(or true seals). The seal baculum is a heterotopic
bone and so it is likely that it shows at least some
growth throughout life. We have found that the size
of the baculum relative to SBL does decrease in old
bulls but perhaps growth layer groups (GLG) can be
determined by histological sectioning of bacula. It
might provide a means to estimate age in very old
individuals where dentition no longer gives useful
estimates of age. Bacular measurements on very old
bulls where the age is known from tagging or from
zoo animals are needed.
Further studies examining the morphology and
growth patterns of the pinniped bacula from known
age animals are required to establish species affinities
and develop identification protocols for seal bacula.
ISS
BACULAR MEASUREMENTS IN SOUTH AFRICAN FUR SEALS
ACKNOWLEDGEMENTS
We wish to express our sincere appreciation to the
following persons and organisations for assistance with
this study: Dr V. Cockcroft (Port Elizabeth Museum), Dr
J. Hanks (WWF-South Africa) and Prof. A. Cockburn
(Australian National University) for financial and logistic
support; Mr. B. Rose (Oosterlig Visserye, Port Elizabeth)
who enabled us to collect seals from his commercial fishing
vessels; staff of the Port Elizabeth Museum for use of bacula
(n = 29) collected before 1992, especially Dr A. Batchelor,
Dr G. Ross and Dr V. Cockcroft; Dr J.H.M David and Mr H.
Oosthuizen (Marine Coastal Management, Cape Town) for
assistance with age determination; Mr N. Minch (Australian
National University) for photographic editing; Dr C. Groves
and Dr A. Thorne (Australian National University) for their
constructive comments on an earlier draft of this manuscript.
This paper is part of a larger study on behalf of the World
Wild Fund For Nature - South Africa (project ZA-348, part
Ic) and a PhD thesis submitted to the Australian National
University in 2001 (Stewardson, 2001).
REFERENCES
Arnould, J.P.Y and Warneke, R.M. (2002). Growth and
condition in Australian fur seals (Arctocephalus
pusillus doriferus) (Carnivora: Pinnepedia).
Australian Journal of Zoology 50, 53-66.
Bester, M.N. (1990). Reproduction in the male Sub
Antarctic fur seal Arctocephalus tropicalis. Journal of
Zoology (London) 222, 177-185.
Burt, W.H. (1939). A study of the baculum in the genera
Perognathus and Dipodomys. Journal of Mammalogy
17, 145-156.
Chane, J. (1925/26). L’Os penien, etude descriptive et
comparative. Actes Société Linnéenne de Bordeaux
78, 5-195.
Cochran, W.G. (1977). Sampling techniques, 3rd Ed.,
(John Wilev and Sons, New York, NY, USA).
Committee on Marine Mammals (1967). Standard
variables of seals. Journal of Mammalogy 48, 459-
462.
David, J.H.M. (1987). South African fur seal
Arctocephalus pusillus pusillus. In ‘Status, biology
and ecology of fur seals’. Proceedings of an
international workshop, Cambridge, England 23-
27 April 1981 (Eds. Croxall, J.P. and Gentry, R.L.)
NOAA Tech. Rep. NMFS 51, 65-71.
David, J.H.M and Rand, R.W. (1986). Attendance
behaviour of South African fur seals. In ‘Fur seals:
Maternal strategies on land and at sea’ (Eds. Gentry,
R.L. and Kooman, G.L.), pp. 126-141. (Princeton
University Press, Princeton, NJ, USA).
Dixson, A.F. (1995). Baculum length and copulatorv
behaviour in carnivores and pinnipeds (Grand Order
Ferae). Journal of Zoology (London) 235, 67-76.
156
Eberhard, W.G. (1985). Sexual selection and animal
genitalia. (Harvard University Press, Cambridge,
MS, USA).
Eberhard, W.G. (1996). Female control: sexual selection
by cryptic female choice. (Princeton University Press,
Princeton, NJ, USA).
Elder, W.H. (1951). The baculum as an age criterion in
mink. Journal of Mammalogy 32, 43-50.
Ewer, R.F. (1973). The carnivores. (Cornell University
Press, New York, NY, USA).
Gibbons, J.D. and Chakraborti, S. (1992). Nonparametric
statistical inference, 3rd Ed., (Marcel Dekker Inc.,
New York, NY, USA).
Griffiths, D., Stirling, W.D. and Weldon, K.L. (1998).
Understanding data. Principles and practice of
statistics. (John Wiley and Sons, New York, NY,
USA).
Hamilton, J.E. (1939). A second report on the Southern
Sea Lion Otaria byronia (de Blainville). Discovery
Report 19, 121-164.
Hewer, J.R. (1964). The determination of age, sexual
maturity and a life-table in the grey seal (Halichoerus
grypus). Proceedings of the Zoological Society of
London 142, 593-624.
Kim, K.C., Repenning, C.A. and Morejohn, G.V. (1975).
Specific antiquity of the sucking lice and evolution
of otariid seals. Rapports et Proces-Verbaux des
Reunions. Conseil International pour |’Exploration de
la Mer 169, 544-549.
King, J.E. (1983). Seals of the World, 2°‘ Ed., (Oxford
University Press, London: British Museum (Nat.
Hist.), London, U.K.).
Laws, R.M. (1956). The elephant seal (Mirounga leonina
Linn.). U1 The physiology of reproduction. Falkland
Islands Dependencies Survey Scientific Reports 15,
1-66.
Laws, R.M. and Sinha, A.A. (1993). Reproduction. In
‘Handbook on Antarctic seal research methods
and techniques’ (Ed. Laws, R.M.), pp. 228-267.
(Cambridge University Press, Cambridge, U.K.).
Lee, M.R. and Schimidly, D.J. (1977). A new species of
Peromyscus (Rodentia: Muridae) from Coahuila,
Mexico. Journal of Mammalogy 58, 263-268.
Long, C.A. and Frank, T. (1968). Morphometric variation
and function in the baculum, with comments on
correlation of parts. Journal of Mammalogy 49, 32-
43.
Mayr, E. (1963). Animal species and evolution. (Belknap
Press, Cambridge, U.K.).
McLaren, I.A. (1960). Are the Pinnipedia biphyletic?
Systematic Zoology 9, 18-28.
Miller, E.H. (2009). Baculum. In “Encyclopedia of marine
mammals’ 2" Ed. (Eds. Perrin, W.F., Wiirsig, B. and
Thewissen, H.). pp 68-71. (Academic Press, San
Diego, CA, USA).
Miller, E.H. (1974). Social behaviour between adult male
and female New Zealand fur seals, Arctocephalus
forsteri (Lesson) during the breeding season.
Australian Journal of Zoology 22, 155-173.
Proc. Linn. Soc. N.S.W., 131, 2010
C.L. STEWARDSON, T. PRVAN AND R.J. RITCHIE
Miller, E.H. and Burton, L.E. (2001). It all relative:
allometry and variation in the baculum (os penis) of
the harp seal, Pagophilus groenlandicus (Carnivora:
Phocidae). Biological Journal of the Linnean Society
72, 345-355.
Miller, E.H., Ponce de Leon, A. and Delong, R.L. (1996).
Violent interspecific sexual behaviour by male
sea lions (Otariidae): evolutionary and phylogenic
implications. Marine Mammal Science 12, 468-476.
Miller, E.H., Stewart, A.R.J. and Stenson, G.B. (1998).
Bacular and testicular growth, allometry, and
variation in the harp seal (Pagophilus groenlandicus).
Journal of Mammalogy 79, 502-513.
Miller, E.H., Jones, I.L. and Stenson, G.B. (1999).
Baculum and testes of the hooded seal (Cystophora
cristata): growth and size-scaling and their
relationships to sexual selection. Canadian Journal of
Zoology 77, 470-470.
Mohr, E. (1963). Os penis und Os clitoridis der
Pinnipedia. Zeitschrift fiir Saugetierkunde 28, 19-37.
Morejohn, G.V. (1975). A phylogeny of otariid seals based
on morphology of the baculum. Rapports et Proces-
Verbaux des Reunions. Conseil International pour
Exploration de la Mer 169, 49-56.
Oosthuizen, W.H. (1997). Evaluation of an effective
method to estimate age of Cape fur seals using
ground tooth sections. Marine Mammal Science 13,
683-693.
Oosthuizen, W.H. and Miller, E,H. (2000). Bacular and
testicular growth and allometry in the Cape fur seal
(Arctocephalus p. pusillus). Marine Mammal Science
16, 124-140.
Patterson, B.D. (1983). Baculum-body size relationships
as evidence for a selective continuum on bacular
morphology. Journal of Mammalogy 64, 496-499.
Patterson, B.D. and Thaeler, C.S. Jr (1982). The
mammalian baculum: hypotheses on the nature of
bacular variability. Journal of Mammalogy 63, 1-15.
Rand, R.W. (1949). Studies on the Cape fur-seal
(Arctocephalus pusillus, Schreber) 3: Age-
grouping in the male. Government Guano Island
Administration, Union of South Africa Department of
Agriculture, Union of South Africa.
Rand, R.W. (1956). The Cape fur seal Arctocephalus
pusillus (Schreber). Its general characteristics and
moult. Sea Fisheries Investigational Report, South
Africa 21, 1-52.
Scheffer, V.B. (1950). Growth of the testes and baculum
in the fur seal, Callorhinus ursinus. Journal of
Mammalogy 31, 384-394.
Scheffer, V.B. and Kenyon, K.W. (1963). Baculum size in
pinnipeds. Zeitschrift fiir Saugetierkunde 28, 38-41.
Stewardson, C.L. (2001). “Biology and conservation of
the Cape (South African) fur seal Arctocephalus
pusillus pusillus (Pinnipedia: Otariidae) from the
Eastern Cape Coast of South Africa’, Thesis (Ph.D.),
Australian National University. http://thesis.anu.edu.
au/public/adt-ANU20030124.162757/Andex.html
Proc. Linn. Soc. N.S.W., 131, 2010
Stewardson, C.L., Bester, M.N. and Oosthuizen, W.H.
(1998). Reproduction in the male Cape fur seal
Arctocephalus pusillus pusillus: age at puberty
and annual cycle of the testis. Journal of Zoology
(London) 246, 63-74.
Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie,
R.J. (2008). Age determination and growth in the
male South African Fur Seal Arctocephalus pusillus
pusillus (Pinnipedia: Otariidae) based upon skull
material. Proceedings of the Linnean Society of NSW
129, 207-252.
Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie,
R.J. (2009). Age determination and growth in the
male South African fur seal Arctocephalus pusillus
pusillus (Pinnipedia: Otariidae) using external body
measurements. Proceedings of the Linnean Society of
NSW 130, 219-244.
Sutton, D.A. and Nadler, C.F. (1974). Systematic revision
of three Townsend chipmunks (Eutamias towns-
endia). Southwestern Naturalist 19, 199-212.
Weisberg, S. (1985). Applied linear regression, 2nd Ed.
(John Wiley and Sons, New York, NY, USA).
Wickens, P.A. (1993). Life expectancy of fur seals with
special reference to the South African (Cape) fur seal.
South African Journal of Wildlife Research 23, 101-
106.
Si
7
1
aon
bafinens. signa. AdatucsaieaeBin soit ial yl Pb rae one
me iy rr ‘i “tains cag a i.e hatha a1 WD ap
nethii aes | Bi con | > a A
By ud Puen lay.am, £ ceca inif ieee rts '%
4 > ) - a.
LHS EA MY AS ST TS 1S SE,
bb RSD PP aD caine sli =a Analige fhet (1 Fag lil eines
Satan ont ni aalederetpe ll, (a toy line Dithorwe si 0 seunethinnta A spade
ee ae wi ee Batons roar) oll “apn hee pag
i p> desl Nea! tes ‘deni Mae Peete adion | ob therwucit tach
sap MOE re AD 4 y oN is ’ cm I
i ' ’ pot ed ae : 1% . ree f > pelieiiee ‘A _sood sh oooh, ua
Hy tells agian ti ois an Olam 0g, yo _wrakys re ccm OM
é Hoe ter Giwt! ee ae vs . deny nme gif Samat) tp poli ATT encil we
thee Leman 9 gig 2) eee) - ’ i ae ah ee bt. LeomaantA-yihebaig aie
‘ acted! hp hn tattle AL lenpeeeD OA) AS Yana Deds A PA ee; aM 7
: iy + eer ae kt tic wn ht aaa 7 :
ea on ek at 2 at 1 rll Yama cee
ie LS ltr thaw 6 iy Ws ste lap ‘
pT cabins ieee Se | Li we ( echin epneuetit,
peatihdd tel Jc gti es 4 — ; sna fii iy Ay i .
i oy tel Ae ' f in) [aa joa oie Vecohe OF ey
- Py Twi!
4
i ats ar Tete a AS, * : qk ws angen fi ievdcatis. be sir ’
Apes ' s ne ' i aot elle et ot ie, ‘a pincer him ts ci
de WW yvummey wens Mt. 1) AN zoo ns
AS dame DN th Herhen, ie iA *sttreonnadinalethl by
Mit ipiaachconeedaionah
fl ssamniame nite nae arie
Pina | ; i) “ite eal ven enlhi
hi Seam anrurryrestit n by
’ i aoe wee i
ne a ” mal ‘anatase ete ie ui
r . balire ect thegrinare
a4 greiner Bes greia . Rana Che: ese Peed
‘ ai F Aca t ber
n \
i i
. se
r
j
stow
4 f }
’ {
L v
vu ' id
{
- " fey,
1
F it a
- =n
‘ eet
cas
eo? mye oan - D> 4
iy
a is ‘ wb’
{ : Pra eserredl
vine Ul Fate, A freimy he Fao
eal" one, 408 wa ‘fet. tite
He
BOOK REVIEW
A Guide To The Beetles Of Australia
George Hangay and Paul Zborowski
CSIRO Publishing
130 Oxford Street ( PO Box 1139)
Collingwood VIC. 3066
RRP :$ 44.95 (Au.)
At last a compact, affordable and beautifully
illustrated beetle book with coloured photos and
easily understood descriptions has been published.
The authors have gone to great lengths to produce
a very comprehensive guide to all the families of
Beetles that occur in Australia and it will be of
great use both by amateur as well as members of
the scientific community. This will enable one to
quickly identify beetles to the family level using the
photos of live beetles and the concise but very clear
descriptions in this new book. The coloured photos
are so superb of the complete adult beetles posing in
their natural habitat, it is like having the live beetle in
ones hand. The photos are often larger in size than the
actual beetle in real life so the identification is often
successful without even the use of a microscope.
The book begins with an Introduction that
explains what makes a beetle and what “ use “ beetles
are, using both photos of world class ( almost all of the
photos, over 400, of the beetles are of live insects ina
natural setting. So this will be so useful for scientific
field or laboratory work, amateur naturalists or even
suburban gardeners wanting identifications.
The next section is on Anatomy and has
very clear descriptions of anatomy, various types of
antennae (the photos of the antennae are clear and
are almost self explanatory to explain the taxonomic
term normally used in the written identification keys).
The thorax, legs and the wings of the beetles have
very clearly drawn illustrations, once again, making
this book more user friendly than so many massive
scientific taxonomic texts of the past.
The next section contains a detailed
description of the Reproduction and Development
of the larvae and adults of Beetles with numerous
photos and clear and very detailed easily understood
descriptions.
Food and Survival section next containing
the type of food that they consume and the defences
they use against their many predators.
Proc. Linn. Soc. N.S.W., 131, 2010
Then, the next section is on the Higher
level of taxonomy of the Beetles, listing suborders,
and all superfamilies. This is a brief outline of four
suborders and a number of superfamilies into which
the families are placed. This also contains both
general and specific details of the superfamilies of
their taxonomic features, feeding habits, ecological
data and information on those superfamilies with
specimens that are known to be pests.
The next section is the Main section of the
book it is the Family Descriptions, this covers all
families that occur in Australia (177 pages). The
Family Descriptions are written in a very clear, concise
and easy to follow with straight forward characters
listed first then photos of world class accuracy of
live beetles, and where no photos they have obtained
permission of illustrations from CSIRO publications.
So all families have at least one form of illustration
be it photo or drawn illustration however some of the
more diverse families have over 20 photos that cover
many of the different genera, within the family.
The Families Descriptions of each Family
covers the following:
1. The known distributions throughout Australia
2. The beetle (larvae and adult) feeding habits
3. The ecology of both the larvae and the adults
4. The numbers of species and genera known in
Australia
5. The photos are a mixture of beetles species from
all states of Australia
6. Common names
After the family descriptions is a list of Endnotes
listing over 53 references, covering taxonomic and
general habits and personal communications from
Australian and overseas scientists who are specialists
studying Beetles.
Next is a very detailed Glossary (over 250 terms
detailed) which explains all the taxonomic or scientific
terms used any where throughout the book.
Next is a very useful Index of “ Common Names
“ of many of the species of Beetles photographed or
59)
described in the book , this is often very useful for
both general and scientific information reporting.
The authors in the process of producing this book
have had direct support by working with over 20 of
Australia and the world’s leading Beetle’s experts on
taxonomic, biology and ecological areas of all the
beetle families of Australia. They have also accessed
other resources such as websites and even photos
of some specimens from the Australian Museum
collections.
As the book is in paperback form and the price
is only $44.95, this allows it to be available to
amateur young insect collectors, naturalists, scientific
laboratories that study insects in details and most
important this makes it a great value for money buy
for University students of Biological and Agriculture
courses. As Beetles are the most commonly found
insects and they are the most prolific insect group
as far as families and species go the lack of a book
covering them has been wanting for many decades.
As Beetles are often the main insect group used in
many environmental impact assessment reports for
the effects over time from suburban or country areas
affected by pollution from such things as mining,
harvesting native forests, pollution from factories etc,
this will be of great benefit in quickly identifying the
beetles for these reports.
I have been assisting to teach Taxonomic
Entomology courses at the University of Sydney
for eleven years now and there has been a desperate
need for a text such as this to give the students the
chance to actually enjoy the learning and studying of
insect identifications. There has been a great need for
160
a book on the taxonomy and details about Beetles of
Australia of this standard for the scientific taxonomic,
biological and ecological University courses for
many years as the only other texts of Beetles prior to
this one has been large scientific volumes covering
all insect groups often costing hundreds of dollars.
As the book is so reasonably priced at $44.95 and is
a compact paperback, the students can take it out on
their collecting field trips to guide them to where and
what beetles they may find as they forage for their
university course collections. The authors have also
added many unusual biological and behaviour notes
that the students will find enjoyable to learn about
beetles as they are the most commonly collected
insect by students. The general public who may be
interested in beetles, will find this a reasonably priced
text, interesting, enjoyable and informative book. The
scientific community will make great use of this new
beetle book as it has the most up to date taxonomic
data, the photos are so clear, concise and so large. The
photos are of beetles from all the different states of
Australia so it will be of great general use throughout
Australia. The easily followed description keys, the
life histories, the food habits, the natural habitat in the
wild, the listing of total species numbers, information
of introduced species, behaviour data, the size range
of the beetles and plant associations all in one book,
make it an absolute must have for all scientific
biological laboratories and institutions working on
any insect research of beetles in Australia.
Elizabeth Jefferys
Sydney
25" May 2010
Proc. Linn. Soc. N.S.W., 131, 2010
INSTRUCTIONS FOR AUTHORS
(this is an abbreviated form — the full instructions can be obtained from our web site or from the Secretary)
1. The Proceedings of the Linnean Society of New South Wales publishes original research papers dealing with
any topic of natural science, particularly biological and earth sciences.
2. Manuscripts should be submitted to the Editor (M.L. Augee, PO Box 82, Kingsford NSW 2032). All
manuscripts are sent to at least two referees and in the first instance three hard copies, including all figures and
tables, must be supplied. Text must be set at one and a half or double spacing.
3. References are cited in the text by the authors’ last name and year of publication (Smith 1987, Smith and
Jones 2000). For three of more authors the citation is (Smith et al. 1988). Notice that commas are not used
between the authors’ names and the year, ‘and’ is spelled out (not &), and et al. is not in italics.
The format for the reference list is:
Journal articles:
Smith, B.S. (1987). A tale of extinction. Journal of Paleontological Fiction 23, 35-78.
Smith, B.S., Wesson, R.I. and Luger, W.K. (1988). Levels of oxygen in the blood of dead Ringtail
Possums. Australian Journal of Sleep 230, 23-53.
Chapters or papers within an edited work:
Ralph, P.H. (2001). The use of ethanol in field studies. In ‘Field techniques’ (Eds. K. Thurstle and P.J.
Green) pp. 34-41. (Northwood Press, Sydney).
Books:
Young, V.H. (1998). ‘The story of the wombat’. (Wallaby Press, Brisbane).
4. An abstract of no more than 200 words is required. Sections in the body of the paper usually include:
INTRODUCTION, MATERIALS AND METHODS, RESULTS, DISCUSSION, ACKNOWLEDGEMENTS
and REFERENCES. Some topics, especially taxonomic, may require variation.
5. Subheadings within the above sections should be in the form:
Bold heading set against left margin
This is the form for the first level headings and the first line of text underneath is indented
Underlined heading set against left margin
This is the next level, and again the first line of text underneath is indented.
Further subheadings should be avoided.
Italics are not to be used for headings but are reserved for genus and species names.
6. Up to 10 KEYWORDS are required. These are often used in computer search engines, so the more specific
the terms the better. ‘Australian’ for example is useless. Please put in alphabetical order.
7. Paragraphs are to be set off by a tab indentation without skipping a line. Do not auto-format the first line (..e.
by using the “first line” command in WORD). All auto-formatting can be fatal when transferring a manuscript
into the publisher platform.
8. Details of setting up the manuscript:
Use 12 point Times New Roman font.
Do not justify
Margins should be: 3 cm top, 2.5 cm bottom, 3 cm left and 2.5 cm right. This is the area available for text;
headers and footers are outside these margins.
9. 3. The final version, incorporating referees’ and editor’s comments, must be supplied on floppy disc or CD in
WORD for PC format (Mac discs will not be accepted). The text file must contain absolutely no autoformatting
or track changes.
INSTRUCTIONS FOR AUTHORS
FIGURES:
Figures can be line drawings, photographs or computer-generated EXCEL or WORD files. No figures
will be accepted larger than 15.5 X 23 cm. Width of lines and sizes of letters in figures must be large enough
to allow reduction to half page size. If a scale is required, it must be presented as a bar within the figure and its
length given in the legend. It is the editor’s prerogative to reduce or enlarge figures as necessary and statements
such as “natural size” or “4X” in the legend are unacceptable.
Photographs must be supplied as black and white prints or as . TIF fi les scanned at 600 dpi. Jpeg is not
acceptable and figures of any kind set in WORD are unacceptable. Line drawings must be supplied as original
copies or as .TIF .files scanned at 1200 dpi. Other figures must be in hardcopy.
While there is no objection to full page size figures, it is journal policy to have the legend on the same
page whenever possible and figures should not be so large as to exclude the legend. Figure legends should be
placed together on a separate page at the end of the manuscript.
TABLES
Because tables may need to be re-sized, it is essential that table legends are not set within the table
but are supplied separately with the figure legends. Legends need to be the same font and size as the rest of the
manuscript.
While the text of the legend is expected to be in 12 point type, it may be necessary to use a smaller font
size for large tables. It is journal policy to accept tables that run over more than one page only in exceptional
circumstances.
Do not use vertical lines in tables unless absolutely necessary to demark data columns. Keep horizontal
lines to a minimum and never put a border around tables.
WORD or EXCEL tables are acceptable, but WORD is preferred.
Tables and/or figures must be separate from the text file. Never embed figures or tables in the text.
10. Details of punctuation, scientific nomenclature, etc. are to be found in the complete instructions available
from the website or from the Secretary.
11. It is helpful if authors suggest a running head of less than 40 characters.
162 Proc. Linn. Soc. N.S.W., 130, 2009
= ; ; pars >
vor: TreeA Mm AUTHORS *”
j ory
1 : i‘ y
* —— = an 4 4 i
‘ a il i
a =i 1G 7 ay ® i ”.
‘peer a , ‘ ¥ f
i '
4 — © i
a La 7
Li re
4 ; i
i a n iy ub
4, y Vv
ate! oi
T i
j ; ; eho h ay
1 aa
ral ve i ia
= ‘s ek
* Ti. ’ Ds
\
i i ‘ \
* =) Ps |
bh mn
i= ip NS + -
ii }, fi
i i ' ee : : :
he Ne : Pea ae -
» hee oh Ga ee ie Peni
" ~ x we - . oo ah
j 7 = v a 7 a r
i x | it a 5 Yi ae i n
“Br, ay
;
17 ad
{
‘ »
: : ‘¥
-
~Y
;
PROCEEDINGS OF THE LINNEAN SOCIETY OF N.S.W.
VOLUME 131
ITUTION LIBRARIES
wi NI
Issued 21 July 2010
CONTENTS
119
141
159
161
Holmes, W.B.K., Anderson, H.M. and Webb, J.A.
The Middle Triassic megafossil flora of the Basin Creek Formation, Nymboida Coal Measures, New South Wales, Australia.
Part 8, The genera Nilssonia, Taeniopteris, Linguifolium, Gontriglossa and Scoresbya.
Cross, D. and Jefferys, E.
Catalogue of insects collected by William Sharp Macleay in Cuba 1825-1836.
Tio, M. and Humphreys, M.
Description of anew species of Inola Davies (Araneae: Pisauridae), the male of |. subtilis Davies and notes on their chromosomes.
Zhen, Y.Y., Burrett, C.F., Percival, |.G. and Lin, BY.
A Late Ordovician conodont fauna from the Lower Limestone Member of the Benjamin Limestone in central Tasmania, and
revision of Tasmanognathus careyi Burrett.
Hunt, J.R. and Young, G.C.
Stratigraphic revision of the Hatchery Creek sequence (Early-Middle Devonian) near Wee Jasper, New South Wales.
Semple, W.S. and Koen, T.B.
Reproductive phenology of white box (Eucalyptus albens Benth.) in the southern portion of its range: 1997-2007.
Sherwin, L. and Meakin, N.S.
The Early Devonian trilobite Craspedarges from the Winduck Group, western New South Wales.
Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie, R.J.
Sexual dimorphism in the adult South African (Cape) fur seal Arctocephalus pulctioe pusillus (Pinnipedia: Otariidae): standard
body length and skull morphology.
Stewardson, C.L., Prvan, T. and Ritchie, R.J. ;
Bacular measurements for age determination and growth in the male South African fur seal Arctocephalus pusillus pusillus
(Pinipedia: Otaridae).
Book Review
Instructions for authors.
Printed by Ligare Pty Ltd,
138 Bonds Rd. Riverwood 2210
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11